[GH-ISSUE #6915 ] Frigate NVR 0.16 #1478

Author

Owner

@MickLesk commented on GitHub (Aug 17, 2025):

In next 2-4 weeks maybe

@MickLesk commented on GitHub (Aug 17, 2025): In next 2-4 weeks maybe

kerem commented

Author

Owner

@iz8mbw commented on GitHub (Aug 17, 2025):

In next 2-4 weeks maybe

Thanks

@iz8mbw commented on GitHub (Aug 17, 2025): > In next 2-4 weeks maybe Thanks

kerem commented

Author

Owner

@xekil commented on GitHub (Aug 17, 2025):

look here -> https://github.com/community-scripts/ProxmoxVE/issues/6912

@xekil commented on GitHub (Aug 17, 2025): look here -> https://github.com/community-scripts/ProxmoxVE/issues/6912

kerem commented

Author

Owner

@iz8mbw commented on GitHub (Aug 17, 2025):

look here -> #6912

Just opened this to track the progress.
Thanks

@iz8mbw commented on GitHub (Aug 17, 2025): > look here -> [#6912](https://github.com/community-scripts/ProxmoxVE/issues/6912) Just opened this to track the progress. Thanks

kerem commented

Author

Owner

@tigertim08 commented on GitHub (Aug 17, 2025):

Thanks

@tigertim08 commented on GitHub (Aug 17, 2025): Thanks

kerem commented

Author

Owner

@MickLesk commented on GitHub (Aug 19, 2025):

There is an first preview, but it doesnt work fine at the moment. The annoying thing about Frigate is that it takes ages to build, and the slightest mistake means you have to start all over again. Unfortunately, that takes up a huge amount of time.

bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVED/main/ct/frigate.sh)"

@MickLesk commented on GitHub (Aug 19, 2025): There is an first preview, but it doesnt work fine at the moment. The annoying thing about Frigate is that it takes ages to build, and the slightest mistake means you have to start all over again. Unfortunately, that takes up a huge amount of time. ```bash bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVED/main/ct/frigate.sh)" ```

kerem commented

Author

Owner

@remz1337 commented on GitHub (Aug 26, 2025):

I will have a look, they may have changed quite a few things in their Dockerfile since v15.

@remz1337 commented on GitHub (Aug 26, 2025): I will have a look, they may have changed quite a few things in their Dockerfile since v15.

kerem commented

Author

Owner

@remz1337 commented on GitHub (Aug 28, 2025):

I have a first draft working (barebone install, no hwaccel, just the sample video playing in loop). Would appreciate if some could test and report any issue. I'll add hwaccel soon

bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/main/ct/frigate.sh)"

@remz1337 commented on GitHub (Aug 28, 2025): I have a first draft working (barebone install, no hwaccel, just the sample video playing in loop). Would appreciate if some could test and report any issue. I'll add hwaccel soon `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/main/ct/frigate.sh)"`

kerem commented

Author

Owner

@MickLesk commented on GitHub (Aug 28, 2025):

Can you Push it into VED? There is the Template

@MickLesk commented on GitHub (Aug 28, 2025): Can you Push it into VED? There is the Template

kerem commented

Author

Owner

@rekenaar00 commented on GitHub (Aug 29, 2025):

I have a first draft working (barebone install, no hwaccel, just the sample video playing in loop). Would appreciate if some could test and report any issue. I'll add hwaccel soon

bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVED/raw/main/ct/frigate.sh)"

Hi, I tried to install on my Proxmox 8.4.9 (13th Gen Intel(R) Core(TM) i7-13620H). It was progressing good up to "downloading model", mabe even two steps further, then a lot of garbage was written to screen, endng with an error. The top part where it started, was not reachable. I copied everyting I could:

diCWH5Z; 7(V0
             ܯW
               J!P&6SV{@ :99Bdds.S/f:ŀHW8
=c.Ժ@^lэ#"HBo\Rϡ=*BdU3Lb}-<z'my0        y Ki0zmYpm|.LP0
/ۉ0T^ٿG8@c%4z][o
@3"5
TInIۏMH&MJ&4Q]Jn4p]S<X1˷7,sP^(Gꁌu$Q6ꮂCG}VX"GT?
                Q)E"#Iҽ1:giάu:k~Y?@($A%.c'&u)!` >xi~q'O`N
                                                         ?M     Ǟswmt7ԹbX$:$XhvN.DQҰCÛJمs3;'#\q9
+<r1=*(`jՆbHs%녠VSN]&4qN
X5SJ)Cu$Mi'MϋupGA8:4*+S
        K3FzG,@^oB@X:ڵl`FVDK
                            "*e,VLFe0c4SQL
q[om    p݁*@\zF-֏֗\`      ]BnauT+"0P^m#wg>X+\Y/B-ĒѨ:'`Q(6{Ƶ=^,3   %6\Cϼu2E"!ն<+UBLm|Fy5$A,Zz^sbC`gH٢.:[C&n輡Gqྵ@iԪjb>lT
                              P 'pS7맂0hiW'8m
*Q~>n,@E<"3QZ2ǹZ1$PJ
                    Eu+Be(Az7o[(yqdO[Q2u(ƩJ
                                           y#+WU#7.__иpr98Ma(ϟٺ`+e}D `7Fmo1Zǎȕ]\v9f\PgN@<7~yؿsX#҅}'FssNs^ٿA<<ԟw}k_@qqxA/_I1{tpR}z(ne
                                             f>#%tVϼuQ~vz)1 41OowznΤ<>U[Eԑ,g*k^X]z.Ì2vaoxNBaEN^63=Y |s        [ϑF3$UvL9-
aMƢ8IVCڝq(S&HUXZ=frgH_f
                      <u2Ӵ҇;ZLToPwF~z@<l{R?UE3ث<591%B[yU-A4e]Peq؅Uu}%:y[ku^Kv+51̵X  L.N
2As"n#qu        Dzn%E   f6eg}:Ci_tư^"a}'O|@w0{ELSBt2*B*Хx)]dpc8Tm 3b~ypxT̡
                                                                        *ƽ{ fȱMǷJ|3b]x1
R{ GCocDڶ<}6C0v!+
Gp˅B<l<qXphOq>܎!VN]q<G  Ϊ0X\N77GYM{pc`XL
                           qgjz?)wa\ܫ*sPE%O1F,Y`hÉ#[ϝX({a(=     bܒ<]w?33d߉6o/tY?*?/~R;PLKrC(.ZLyRІU̻TSX栜A|7 'I^h<
Hn$)ƤQFr:Ar(:LPk%u\ksn.UWS[
e)җzUt'y/13obMuэL7
                  \6َLwPRtÔ[Gw
hp
  /
   s*!r
qԽ#RiGguEQIn1Tdr9L%DPy~r,5$
                           g+ߡ 3i]kefh6=:6S/!$x~?3Zg4dP&Y\G)zSDOdYJ\ɔ]UgvXLV4%BSv%Z-Oa#aD7#C~O|/z3Z"}z3
                 A}'    cAwFdH+}4"ÖHmAulDx~iMގwt;rwY
                                                    }k4p.o'0(G}biχ:T[Up<
                                                                         QQ8G(+1p-I`
9xh8`090
lF|#RyD>X8s*8#{!f\pU>F}5"O#b"K
m+뻔7;K+aǭ0dϹWָ[7ZhܯzA0Z-ࡊaqr(.=Ƹz+Y8mMb$,@ЛoPUs{~m
                                                        7!Ӹgd=:xs2sĵ*;/vA~|˿~SJdHߗ\s7t#'h`{}߷\
        'Yz9;LYdQ_Ɋ[Ybպ)d=L&h4I*Ya<l-3wgo%PiiV59W=37&!IZ1
                                                         "Xskse/yIW     Yr3j
                                                                            g՛}Y$fVVDO̗yGֽ6EI}Bpgi,RIKoK{;HdƑLŉ!c"3a'ԙAdlGv5Ur/F3RYԐűhr"#xك,<d: {W嵅E6- EY<.EPpuGYTDbYT+VY,bUDlBE܀Bi/yڼ%o$'?ܛs7gVFK<z
f[Ĥ/caǬٖ!@Әysհ!j>       ЛŸ0قՎfRj%CS)
e`:1;yYElw72k"Yv:
                YIp?!/b 범YB%;uEUk=/E>_4//S_@   %
Xړ
  *m2   xmcU    IS}lL+ R<2eARs  /\9\M)4cqרO[֓
                                            xДz85>#!
}P眀0G[|W?Z͘
           ҇GOw_IT

                  q~A+fiJBuf7o@ ^V
u?G{MHacV 0*|J"f?T
                  8/.zV܈¾p)K
                            {'߂
                               3{#G𱳵P52Eg31k?@AX|Ҵx?#'pEP<uCt2_rr VUJ]ۀU:qWTpM^oQ
                                                                                  ^Ԍ济}^*g7<FP2
QoGU㹯Q['Bqb^:  NnwW]Rˇ4/_}Sw>܇43Ӈ4}V4YQ*ѝH ^.
                                              qד1BkrLnQD$ȗR,*gice'HZBuXcwa@)qT  XգZ)&&:|G?<K'fPJQaE[I4>ҭijl9DҌnf$KH*@τ4I:)ͤ+ɪ%<~"3<u8˲BuVswwWdN
                                                        ag_"ϵ٪cG2@$u-?<F.R{A3Ǒr>}Cӧ.mh,)l~@<bI_._.GEVL^&<rr쪡Ѕ&_LErA}9Һ>АN*S(Udڳܟ
                                           fkGg&ãTmwdpr<#o褫2l4^Picݪ1s5bŧ^jyڞg*;!My苗֯>UH bxRZp6H4nK^leF0]ux
kТabĺh:_vFdX3,8vb<_FfZg=[?_T9O*َ7[Pꄔ>!m44Ɩבp< =ℕIB|;ã}0Rt=^
                        )N[۬Ci!
.rxIz1C4yCۺ"'Kj(?i      a!9O)8+6^TbE0MgOYN-NHvIO̾<j%A/v;VAllPyfx46#4s[Dw)-YVFtu
-}$o                                                                           ]
+)',if*_#8n>Bs\+T
         gd7Q3umr,0slK/TN/$oyYO"gMHk (TOoëd,IƑѰ;gO 50+Ƨ%<W)rA\p/Q8<r;cY%zdd :A\*μ/*h*6m
 f]\'3p-aZ)
^}l9Tg"1?G6r~JwxZf>@-+6[W;߷?%U3o|9w't|I}&+$             `!a|2kLm=X_ڨp,MHBEd7>mP=
                                                                                W%$|@Ym|yXksRDe?Fyq@9ņx*P`hvwZ5ܪԏmڗIWiHMԫy4UwIN
                                         /%lOY`JB&{ZL.»4c噍f7֨EIџ2^E$SIO7Հ%wͿs1rܪY$肅(-sQ6>3X0y*vbuwOxcsHq/?G5Uݑ|ZbڳBP0*;qv^ix{746`\c075_6,$T
                                                      `M ܆X!rr`+t!CBPGF ~~Zt!^ƮѠZ    X_םgIw-
./S*ǼJm(6V􁎸.
Q.lջ{yjcc+m%樯!]{!-a6]Չ$%NYDUN3zfno^*\2
                                       H)\&Ta;['%bURP
ݳ;wH`v{{@:ҲSvu^;mןq[~:_]Cל
                           .9/L;7t<K[׹ߝ3}c?ts@~|
                                                >{2Zl*wjVzX哏Axxs_S>4S->o*?pKw}g|Ӎ.SwKNoTB[W{O'/^V^{y?/+V>Y柜Q4/?xW˯\t[y-n9_Dg}]E/Z;۝)"zKRkO0~>+c'zxs͑sgur|!~=c4ϝgڷ
                                                                           mڞد?|y
                                                                                 ^W;;]lkͯ<~]tVsΡ/й\|v:kxu߰.3Ԏt}>Zb걏]|m!1sp蕧.x/J8Vz[&};'J}<]/
                                                      }o*8toiSxK?Ћ+J.Psogvw&^:8~zߟV>2y{n//=u(:Tu z{ݺr}^/
kYAv1yI`?s{OsA?œsvi~<Ṿl$n*dAϲ-KQޞ%ġ?g=z^7Nbq+x7,FD=)?һk-       ǭv{2X2O%?zf,LłSNr
                               l.+1[{{6Qڛ(A
3_+90>Fp4"?<?y&y鯴$i-Zzb

hkhc}R&#zbg7            5nRYUuRkv®,{E؞.eǺݥYL*:@sǖWXuc&;'W?
            1|3d~O}$0g
6>=ipA'ez=:fX[yRi     5dYGU=n{_C`g|и5Ki~̖93>_usx2= 
                 mD'AHF7cpT
Z`}'>(2|>yZ/8XJ˕/世xym\
                       \Ff.4ֽ_^
                              .Whm9^C8[!Cc
kcCɀ{U|\&,QhmgRрDa$qkU>m"Ҫ4,G-kEu&䡝J4KncYhm?zV%n^y
                                                   ę6"bF;UB0=`zD2Ik     B%F$uCq_OhMC\vfgE>nge1}g:ޘ$c2
:jPiC:\D0ƒؚ$zBtw3 QYSod?]4Iy[X.ZEIɺVd
                                     YPL
M5ȸp,FXGUX¥3f=$%714prJ(UYuبF"Y,w

W([#Q6Ss042YlZ7Z5ִ`PQ7Hvay'3)   M"Ih|k%i3tiQ\(I:/5aimBV9EnMDTުLCMk
tJ|#<G5C֝>)sa7pT@Iz      S"b*-a7$R~$d                             }br{הQ(%?=̭ak(|N_t&
                                    -q/&AJtz88
                                              Y[
"5>|jI®uiu5H,x:f>Iѥ3Gn`TYof>v@Jcae2V@#j;R               P!WG)׍lFJ7> b
PJ8kh0*^Wj{                              ,@̬RL,}|31gMrfIB
           PjzF К#
                  ^'5^!'v
R*GxP8eiD
         !mHS/Eܳys>}]Ԩr3%eIa[
)ֹ-uN\FO20Wpa4*z,b;#5r3G*γ;b-"pn>BbaC5yW
jʬJJTesΰNFd     ?nBi(Sfq     Pء@KYLFL6C lͫd{8kOacY؄1RdEޯCt1a#s\+@86ɤ
                        C|      9$F3&I*D
4                                      d7Uʱ+2foC 
D$fKBp>!E2 U Ѷְzrg=d6Ⱥ6|EXؐ(
                          X_ކD 0#ahMdU  rC<2D`uuIS.Soґ"9:F-ém8#:<(| n<r#gsX     bTɦyWU8ú@3&Fq+^
         䪜)+u)3J,%)Id\H5dkUgڀE;2S
I}QFQыkVX9VA{PU F*6P;N>S.B&kٍ2գ!B79RGťiNؤA"~
L(E"!EoSkCzP$Ղ4ũ1Tx8    ^oBY3D8n^t2KfOq91_!ٜ5f#R        Udbm=%Sʐ|^O*TcχœmuA#W6eU/[\ K-
                                                                                     q-3DN IB&
5[#6@TS>-<23H(''st_/)fTJMՀnat@mssj0&@}IA+CGK}b.ւ#7f.\!g
                                                       ڜȏ,A"a4zE#˃D@GfÓGՍC>F7]6n$SSN`=;uy]99  *?k~4ThT%
    Z
     D\>xQ'T]#Qۊlwe!
I@Pg^bM՝xK
եFA0fؠ40ʲ3Rc^
Ցt5v6Bq(<^CuTkQkDMj~5aXE'3*t̘hM'L@B      (1]|IU;MX:TT[)>hTAQUTѰ`OPz
hؽ      ҉Rie
            ЪNC-cUcYHϐ6!?Q|!)Upހ        ݞ$jR۾
                                             E9flmY/(2PѪľ9TTRO")>ݠT6T+8@*)3x
+]TH\^TQ2D5O@#SOX7yZ                                                        £3#_X|.>&
Ps E)[MdJ=EI8I@bi0iEJ$)6=L&,;U@Q\QgƸ    Q
    R"$bT錵V̂BMr\#ef_S%eyXk-x׻ϟeaۯ{&R5
B៻NRJojcj8ە9]54"U7Pz! ^rj            ]
üb?";<M$WЄ8BM

#2Ｕoץ%ԛ[˝oq'|3ck܂ϫn4a3
                       cNASbu-6Lpޅ5g*KnUagm6@hP+G珯֎Fd^ !!!!!_~@4 ?
gzip: stdin: unexpected end of file
tar: Child returned status 1
tar: Error is not recoverable: exiting now
⠏ Building Models
[ERROR] in line 136: exit code 0: while executing command tar xzf yamnet-tflite-classification-tflite-v1.tar.gz

[ERROR] in line 1446: exit code 2: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)"

/dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required

@rekenaar00 commented on GitHub (Aug 29, 2025): > I have a first draft working (barebone install, no hwaccel, just the sample video playing in loop). Would appreciate if some could test and report any issue. I'll add hwaccel soon > > `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVED/raw/main/ct/frigate.sh)"` Hi, I tried to install on my Proxmox 8.4.9 (13th Gen Intel(R) Core(TM) i7-13620H). It was progressing good up to "downloading model", mabe even two steps further, then a lot of garbage was written to screen, endng with an error. The top part where it started, was not reachable. I copied everyting I could: ``` diCWH5Z; 7(V0 ܯW J!P&6SV{@ :99Bdds.S/f:ŀHW8 =c.Ժ@^lэ#"HBo\Rϡ=*BdU3Lb}-<z'my0 y Ki0zmYpm|.LP0 /ۉ0T^ٿG8@c%4z][o @3"5 TInIۏMH&MJ&4Q]Jn4p]S<X1˷7,sP^(Gꁌu$Q6ꮂCG}VX"GT? Q)E"#Iҽ1:giάu:k~Y?@($A%.c'&u)!` >xi~q'O`N ?M Ǟswmt7ԹbX$:$XhvN.DQҰCÛJمs3;'#\q9 +<r1=*(`jՆbHs%녠VSN]&4qN X5SJ)Cu$Mi'MϋupGA8:4*+S K3FzG,@^oB@X:ڵl`FVDK "*e,VLFe0c4SQL q[om p݁*@\zF-֏֗\` ]BnauT+"0P^m#wg>X+\Y/B-ĒѨ:'`Q(6{Ƶ=^,3 %6\Cϼu2E"!ն<+UBLm|Fy5$A,Zz^sbC`gH٢.:[C&n輡Gqྵ@iԪjb>lT P 'pS7맂0hiW'8m *Q~>n,@E<"3QZ2ǹZ1$PJ Eu+Be(Az7o[(yqdO[Q2u(ƩJ y#+WU#7.__иpr98Ma(ϟٺ`+e}D `7Fmo1Zǎȕ]\v9f\PgN@<7~yؿsX#҅}'FssNs^ٿA<<ԟw}k_@qqxA/_I1{tpR}z(ne f>#%tVϼuQ~vz)1 41OowznΤ<>U[Eԑ,g*k^X]z.Ì2vaoxNBaEN^63=Y |s [ϑF3$UvL9- aMƢ8IVCڝq(S&HUXZ=frgH_f <u2Ӵ҇;ZLToPwF~z@<l{R?UE3ث<591%B[yU-A4e]Peq؅Uu}%:y[ku^Kv+51̵X L.N 2As"n#qu Dzn%E f6eg}:Ci_tư^"a}'O|@w0{ELSBt2*B*Хx)]dpc8Tm 3b~ypxT̡ *ƽ{ fȱMǷJ|3b]x1 R{ GCocDڶ<}6C0v!+ Gp˅B<l<qXphOq>܎!VN]q<G Ϊ0X\N77GYM{pc`XL qgjz?)wa\ܫ*sPE%O1F,Y`hÉ#[ϝX({a(= bܒ<]w?33d߉6o/tY?*?/~R;PLKrC(.ZLyRІU̻TSX栜A|7 'I^h< Hn$)ƤQFr:Ar(:LPk%u\ksn.UWS[ e)җzUt'y/13obMuэL7 \6َLwPRtÔ[Gw hp / s*!r qԽ#RiGguEQIn1Tdr9L%DPy~r,5$ g+ߡ 3i]kefh6=:6S/!$x~?3Zg4dP&Y\G)zSDOdYJ\ɔ]UgvXLV4%BSv%Z-Oa#aD7#C~O|/z3Z"}z3 A}' cAwFdH+}4"ÖHmAulDx~iMގwt;rwY }k4p.o'0(G}biχ:T[Up< QQ8G(+1p-I` 9xh8`090 lF|#RyD>X8s*8#{!f\pU>F}5"O#b"K m+뻔7;K+aǭ0dϹWָ[7ZhܯzA0Z-ࡊaqr(.=Ƹz+Y8mMb$,@ЛoPUs{~m 7!Ӹgd=:xs2sĵ*;/vA~|˿~SJdHߗ\s7t#'h`{}߷\ 'Yz9;LYdQ_Ɋ[Ybպ)d=L&h4I*Ya<l-3wgo%PiiV59W=37&!IZ1 "Xskse/yIW Yr3j g՛}Y$fVVDO̗yGֽ6EI}Bpgi,RIKoK{;HdƑLŉ!c"3a'ԙAdlGv5Ur/F3RYԐűhr"#xك,<d: {W嵅E6- EY<.EPpuGYTDbYT+VY,bUDlBE܀Bi/yڼ%o$'?ܛs7gVFK<z f[Ĥ/caǬٖ!@Әysհ!j> ЛŸ0قՎfRj%CS) e`:1;yYElw72k"Yv: YIp?!/b 범YB%;uEUk=/E>_4//S_@ % Xړ *m2 xmcU IS}lL+ R<2eARs /\9\M)4cqרO[֓ xДz85>#! }P眀0G[|W?Z͘ ҇GOw_IT q~A+fiJBuf7o@ ^V u?G{MHacV 0*|J"f?T 8/.zV܈¾p)K {'߂ 3{#G𱳵P52Eg31k?@AX|Ҵx?#'pEP<uCt2_rr VUJ]ۀU:qWTpM^oQ ^Ԍ济}^*g7<FP2 QoGU㹯Q['Bqb^: NnwW]Rˇ4/_}Sw>܇43Ӈ4}V4YQ*ѝH ^. qד1BkrLnQD$ȗR,*gice'HZBuXcwa@)qT XգZ)&&:|G?<K'fPJQaE[I4>ҭijl9DҌnf$KH*@τ4I:)ͤ+ɪ%<~"3<u8˲BuVswwWdN ag_"ϵ٪cG2@$u-?<F.R{A3Ǒr>}Cӧ.mh,)l~@<bI_._.GEVL^&<rr쪡Ѕ&_LErA}9Һ>АN*S(Udڳܟ fkGg&ãTmwdpr<#o褫2l4^Picݪ1s5bŧ^jyڞg*;!My苗֯>UH bxRZp6H4nK^leF0]ux kТabĺh:_vFdX3,8vb<_FfZg=[?_T9O*َ7[Pꄔ>!m44Ɩבp< =ℕIB|;ã}0Rt=^ )N[۬Ci! .rxIz1C4yCۺ"'Kj(?i a!9O)8+6^TbE0MgOYN-NHvIO̾<j%A/v;VAllPyfx46#4s[Dw)-YVFtu -}$o ] +)',if*_#8n>Bs\+T gd7Q3umr,0slK/TN/$oyYO"gMHk (TOoëd,IƑѰ;gO 50+Ƨ%<W)rA\p/Q8<r;cY%zdd :A\*μ/*h*6m f]\'3p-aZ) ^}l9Tg"1?G6r~JwxZf>@-+6[W;߷?%U3o|9w't|I}&+$ `!a|2kLm=X_ڨp,MHBEd7>mP= W%$|@Ym|yXksRDe?Fyq@9ņx*P`hvwZ5ܪԏmڗIWiHMԫy4UwIN /%lOY`JB&{ZL.»4c噍f7֨EIџ2^E$SIO7Հ%wͿs1rܪY$肅(-sQ6>3X0y*vbuwOxcsHq/?G5Uݑ|ZbڳBP0*;qv^ix{746`\c075_6,$T `M ܆X!rr`+t!CBPGF ~~Zt!^ƮѠZ X_םgIw- ./S*ǼJm(6V􁎸. Q.lջ{yjcc+m%樯!]{!-a6]Չ$%NYDUN3zfno^*\2 H)\&Ta;['%bURP ݳ;wH`v{{@:ҲSvu^;mןq[~:_]Cל .9/L;7t<K[׹ߝ3}c?ts@~| >{2Zl*wjVzX哏Axxs_S>4S->o*?pKw}g|Ӎ.SwKNoTB[W{O'/^V^{y?/+V>Y柜Q4/?xW˯\t[y-n9_Dg}]E/Z;۝)"zKRkO0~>+c'zxs͑sgur|!~=c4ϝgڷ mڞد?|y ^W;;]lkͯ<~]tVsΡ/й\|v:kxu߰.3Ԏt}>Zb걏]|m!1sp蕧.x/J8Vz[&};'J}<]/ }o*8toiSxK?Ћ+J.Psogvw&^:8~zߟV>2y{n//=u(:Tu z{ݺr}^/ kYAv1yI`?s{OsA?œsvi~<Ṿl$n*dAϲ-KQޞ%ġ?g=z^7Nbq+x7,FD=)?һk- ǭv{2X2O%?zf,LłSNr l.+1[{{6Qڛ(A 3_+90>Fp4"?<?y&y鯴$i-Zzb hkhc}R&#zbg7 5nRYUuRkv®,{E؞.eǺݥYL*:@sǖWXuc&;'W? 1|3d~O}$0g 6>=ipA'ez=:fX[yRi 5dYGU=n{_C`g|и5Ki~̖93>_usx2= mD'AHF7cpT Z`}'>(2|>yZ/8XJ˕/世xym\ \Ff.4ֽ_^ .Whm9^C8[!Cc kcCɀ{U|\&,QhmgRрDa$qkU>m"Ҫ4,G-kEu&䡝J4KncYhm?zV%n^y ę6"bF;UB0=`zD2Ik B%F$uCq_OhMC\vfgE>nge1}g:ޘ$c2 :jPiC:\D0ƒؚ$zBtw3 QYSod?]4Iy[X.ZEIɺVd YPL M5ȸp,FXGUX¥3f=$%714prJ(UYuبF"Y,w W([#Q6Ss042YlZ7Z5ִ`PQ7Hvay'3) M"Ih|k%i3tiQ\(I:/5aimBV9EnMDTުLCMk tJ|#<G5C֝>)sa7pT@Iz S"b*-a7$R~$d }br{הQ(%?=̭ak(|N_t& -q/&AJtz88 Y[ "5>|jI®uiu5H,x:f>Iѥ3Gn`TYof>v@Jcae2V@#j;R P!WG)׍lFJ7> b PJ8kh0*^Wj{ ,@̬RL,}|31gMrfIB PjzF К# ^'5^!'v R*GxP8eiD !mHS/Eܳys>}]Ԩr3%eIa[ )ֹ-uN\FO20Wpa4*z,b;#5r3G*γ;b-"pn>BbaC5yW jʬJJTesΰNFd ?nBi(Sfq Pء@KYLFL6C lͫd{8kOacY؄1RdEޯCt1a#s\+@86ɤ C| 9$F3&I*D 4 d7Uʱ+2foC D$fKBp>!E2 U Ѷְzrg=d6Ⱥ6|EXؐ( X_ކD 0#ahMdU rC<2D`uuIS.Soґ"9:F-ém8#:<(| n<r#gsX bTɦyWU8ú@3&Fq+^ 䪜)+u)3J,%)Id\H5dkUgڀE;2S I}QFQыkVX9VA{PU F*6P;N>S.B&kٍ2գ!B79RGťiNؤA"~ L(E"!EoSkCzP$Ղ4ũ1Tx8 ^oBY3D8n^t2KfOq91_!ٜ5f#R Udbm=%Sʐ|^O*TcχœmuA#W6eU/[\ K- q-3DN IB& 5[#6@TS>-<23H(''st_/)fTJMՀnat@mssj0&@}IA+CGK}b.ւ#7f.\!g ڜȏ,A"a4zE#˃D@GfÓGՍC>F7]6n$SSN`=;uy]99 *?k~4ThT% Z D\>xQ'T]#Qۊlwe! I@Pg^bM՝xK եFA0fؠ40ʲ3Rc^ Ցt5v6Bq(<^CuTkQkDMj~5aXE'3*t̘hM'L@B (1]|IU;MX:TT[)>hTAQUTѰ`OPz hؽ ҉Rie ЪNC-cUcYHϐ6!?Q|!)Upހ ݞ$jR۾ E9flmY/(2PѪľ9TTRO")>ݠT6T+8@*)3x +]TH\^TQ2D5O@#SOX7yZ £3#_X|.>& Ps E)[MdJ=EI8I@bi0iEJ$)6=L&,;U@Q\QgƸ Q R"$bT錵V̂BMr\#ef_S%eyXk-x׻ϟeaۯ{&R5 B៻NRJojcj8ە9]54"U7Pz! ^rj ] üb?";<M$WЄ8BM #2Ｕoץ%ԛ[˝oq'|3ck܂ϫn4a3 cNASbu-6Lpޅ5g*KnUagm6@hP+G珯֎Fd^ !!!!!_~@4 ? gzip: stdin: unexpected end of file tar: Child returned status 1 tar: Error is not recoverable: exiting now ⠏ Building Models [ERROR] in line 136: exit code 0: while executing command tar xzf yamnet-tflite-classification-tflite-v1.tar.gz [ERROR] in line 1446: exit code 2: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)" /dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required ```

kerem commented

Author

Owner

@remz1337 commented on GitHub (Aug 30, 2025):

@rekenaar00 should be fixed now, please try again

@remz1337 commented on GitHub (Aug 30, 2025): @rekenaar00 should be fixed now, please try again

kerem commented

Author

Owner

@rekenaar00 commented on GitHub (Aug 30, 2025):

@rekenaar00 should be fixed now, please try again

yes, it worked now, thanks.
I will test further when hwaccel added.

@rekenaar00 commented on GitHub (Aug 30, 2025): > [@rekenaar00](https://github.com/rekenaar00) should be fixed now, please try again yes, it worked now, thanks. I will test further when hwaccel added.

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 1, 2025):

I've added support for hwaccel. @rekenaar00 can you test again please?

@remz1337 commented on GitHub (Sep 1, 2025): I've added support for hwaccel. @rekenaar00 can you test again please?

kerem commented

Author

Owner

@iz8mbw commented on GitHub (Sep 1, 2025):

Hello all.
I'm confused if I should test/use the script provided by @MickLesk or the one provided by @remz1337.
What is the difference? What is/will be the "official" one of community-scripts?

Thanks.

@iz8mbw commented on GitHub (Sep 1, 2025): Hello all. I'm confused if I should test/use the script provided by @MickLesk or the one provided by @remz1337. What is the difference? What is/will be the "official" one of community-scripts? Thanks.

kerem commented

Author

Owner

@MickLesk commented on GitHub (Sep 1, 2025):

It can be merged into ProxmoxVED at any time. That would also be the cleanest way to find out whether it can then be migrated to Live (ProxmoxVE).

@MickLesk commented on GitHub (Sep 1, 2025): It can be merged into ProxmoxVED at any time. That would also be the cleanest way to find out whether it can then be migrated to Live (ProxmoxVE).

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 1, 2025):

@remz1337 , I just gave it a go, this is the bottom of the output :

Setting up libigdfcl1:amd64 (1.0.12504.6-1+deb12u1) ...
Setting up intel-opencl-icd (22.43.24595.41-1) ...
Processing triggers for libc-bin (2.36-9+deb12u10) ...
Processing triggers for man-db (2.11.2-2) ...
chgrp: cannot access '/dev/dri': No such file or directory

[ERROR] in line 24: exit code 0: while executing command chgrp video /dev/dri

[ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)"

/dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required
root@pve:~#

@Zixim commented on GitHub (Sep 1, 2025): @remz1337 , I just gave it a go, this is the bottom of the output : ``` Setting up libigdfcl1:amd64 (1.0.12504.6-1+deb12u1) ... Setting up intel-opencl-icd (22.43.24595.41-1) ... Processing triggers for libc-bin (2.36-9+deb12u10) ... Processing triggers for man-db (2.11.2-2) ... chgrp: cannot access '/dev/dri': No such file or directory [ERROR] in line 24: exit code 0: while executing command chgrp video /dev/dri [ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)" /dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required root@pve:~# ```

kerem commented

Author

Owner

@iz8mbw commented on GitHub (Sep 1, 2025):

@remz1337 , I just gave it a go, this is the bottom of the output :

Setting up libigdfcl1:amd64 (1.0.12504.6-1+deb12u1) ...
Setting up intel-opencl-icd (22.43.24595.41-1) ...
Processing triggers for libc-bin (2.36-9+deb12u10) ...
Processing triggers for man-db (2.11.2-2) ...
chgrp: cannot access '/dev/dri': No such file or directory

[ERROR] in line 24: exit code 0: while executing command chgrp video /dev/dri

[ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)"

/dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required
root@pve:~#

I have the same error using the remz1337 script:

chgrp: cannot access '/dev/dri': No such file or directory

[ERROR] in line 24: exit code 0: while executing command chgrp video /dev/dri

[ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)"

/dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required
root@pve01:~#

but I still have the "old" frigate LXC UP&Running.
May be I need to remove the old LXC to freeup the "video"?

@iz8mbw commented on GitHub (Sep 1, 2025): > [@remz1337](https://github.com/remz1337) , I just gave it a go, this is the bottom of the output : > > ``` > Setting up libigdfcl1:amd64 (1.0.12504.6-1+deb12u1) ... > Setting up intel-opencl-icd (22.43.24595.41-1) ... > Processing triggers for libc-bin (2.36-9+deb12u10) ... > Processing triggers for man-db (2.11.2-2) ... > chgrp: cannot access '/dev/dri': No such file or directory > > [ERROR] in line 24: exit code 0: while executing command chgrp video /dev/dri > > [ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)" > > /dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required > root@pve:~# > ``` I have the same error using the remz1337 script: ``` chgrp: cannot access '/dev/dri': No such file or directory [ERROR] in line 24: exit code 0: while executing command chgrp video /dev/dri [ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)" /dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required root@pve01:~# ``` but I still have the "old" frigate LXC UP&Running. May be I need to remove the old LXC to freeup the "video"?

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 1, 2025):

In my case, there is no other Frigate instance, there is a Plex lxc which also has HW acceleration.
However, this shouldn't be any issue because hw acceleration is shared across lxc.

edited : typo

@Zixim commented on GitHub (Sep 1, 2025): In my case, there is no other Frigate instance, there _is_ a Plex lxc which also has HW acceleration. However, this shouldn't be any issue because hw acceleration is **shared** across lxc. edited : typo

kerem commented

Author

Owner

@ShaMAD commented on GitHub (Sep 1, 2025):

I've added support for hwaccel. @rekenaar00 can you test again please?

Hi, some issues during install on first Proxmox:

✔️ Network in LXC is reachable (ping)
⏳ Customizing LXC Containersed: -e expression #1, char 0: no previous regular expression

[ERROR] in line 1423: exit code 1: while executing command pct exec "$CTID" -- bash -c "sed -i '/$LANG/ s/^# //' /etc/locale.gen"

/dev/fd/63: line 49: exit: pct exec "$CTID" -- bash -c "sed -i '/$LANG/ s/^# //' /etc/locale.gen": numeric argument required

On another Proxmox server:
Processing triggers for man-db (2.11.2-2) ...
chgrp: cannot access '/dev/dri': No such file or directory

[ERROR] in line 24: exit code 0: while executing command chgrp video /dev/dri

[ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)"

/dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required

@ShaMAD commented on GitHub (Sep 1, 2025): > I've added support for hwaccel. [@rekenaar00](https://github.com/rekenaar00) can you test again please? Hi, some issues during install on first Proxmox: ✔️ Network in LXC is reachable (ping) ⏳ Customizing LXC Containersed: -e expression #1, char 0: no previous regular expression [ERROR] in line 1423: exit code 1: while executing command pct exec "$CTID" -- bash -c "sed -i '/$LANG/ s/^# //' /etc/locale.gen" /dev/fd/63: line 49: exit: pct exec "$CTID" -- bash -c "sed -i '/$LANG/ s/^# //' /etc/locale.gen": numeric argument required On another Proxmox server: Processing triggers for man-db (2.11.2-2) ... chgrp: cannot access '/dev/dri': No such file or directory [ERROR] in line 24: exit code 0: while executing command chgrp video /dev/dri [ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)" /dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 1, 2025):

lxc.mount.entry: /dev/dri/card0 /dev/dri/card0 none bind,optional,create=file is mounted in the frigate lxc,
but /dev/dri isn't.

Hence : chgrp: cannot access '/dev/dri': No such file or directory

edit : I see now that HW-acceleration was added only for Nvidia. Then it makes sense to mount /dev/dri/card0

@Zixim commented on GitHub (Sep 1, 2025): `lxc.mount.entry: /dev/dri/card0 /dev/dri/card0 none bind,optional,create=file` is mounted in the frigate lxc, but `/dev/dri` isn't. Hence : `chgrp: cannot access '/dev/dri': No such file or directory` edit : I see now that HW-acceleration was added only for Nvidia. Then it makes sense to mount `/dev/dri/card0`

kerem commented

Author

Owner

@drshockz commented on GitHub (Sep 1, 2025):

Hey, does anyone know why when I install It works all good I can load webui, but then all it shows is a spinning icon. and Im logged in as user anonymous (undefined), and none of the /api routes like /api/logout work.

@drshockz commented on GitHub (Sep 1, 2025): Hey, does anyone know why when I install It works all good I can load webui, but then all it shows is a spinning icon. and Im logged in as user anonymous (undefined), and none of the /api routes like /api/logout work.

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 1, 2025):

@drshockz Because this (frigate v0.16) is under development, not even alfa or beta.

@Zixim commented on GitHub (Sep 1, 2025): @drshockz Because this (frigate v0.16) is under development, not even alfa or beta.

kerem commented

Author

Owner

@drshockz commented on GitHub (Sep 1, 2025):

I see, so like I shouldn't really expect this to work at all lol. What do you think is causing that, or is that like where you guys are at?

@drshockz commented on GitHub (Sep 1, 2025): I see, so like I shouldn't really expect this to work at all lol. What do you think is causing that, or is that like where you guys are at?

kerem commented

Author

Owner

@ams2990 commented on GitHub (Sep 1, 2025):

You shouldn't expect it to work at all. Not even all the testers have gotten it to install fully. Be patient.

@ams2990 commented on GitHub (Sep 1, 2025): You shouldn't expect it to work at all. Not even all the testers have gotten it to install fully. Be patient.

kerem commented

Author

Owner

@Telemaniaka commented on GitHub (Sep 1, 2025):

I am not knowledgeble enough to say if this is indeed what fixed the hw acceleration for me or why,

But they showed up after I removed the / slash infront of the target for the /dev/dri mount commands

I saw it from the old working frigate 14 instance

lxc.mount.entry: /dev/dri dev/dri none bind,optional,create=dir
lxc.mount.entry: /dev/dri/renderD128 dev/dri/renderD128 none bind,optional,create=file

(I'm using the integrated gpu, exact file names may vary for you)

@Telemaniaka commented on GitHub (Sep 1, 2025): I am not knowledgeble enough to say if this is indeed what fixed the hw acceleration for me or why, But they showed up after I removed the `/` slash infront of the target for the `/dev/dri` mount commands I saw it from the old working frigate 14 instance ``` lxc.mount.entry: /dev/dri dev/dri none bind,optional,create=dir lxc.mount.entry: /dev/dri/renderD128 dev/dri/renderD128 none bind,optional,create=file ``` (I'm using the integrated gpu, exact file names may vary for you)

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 1, 2025):

thanks all for testing, I'll have a look

@remz1337 commented on GitHub (Sep 1, 2025): thanks all for testing, I'll have a look

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 1, 2025):

@Telemaniaka When/where did you do this? I know it is the LXC conf file in /etc/pve/lxc but it gets created by the helper script, the LXC gets started with the original configuration then eventually fails. It seems like it needs to be edited before the script is run, but you can't do that because then the script thinks the LXC already exists. Thanks.

@atrozgrifo commented on GitHub (Sep 1, 2025): @Telemaniaka When/where did you do this? I know it is the LXC conf file in /etc/pve/lxc but it gets created by the helper script, the LXC gets started with the original configuration then eventually fails. It seems like it needs to be edited before the script is run, but you can't do that because then the script thinks the LXC already exists. Thanks.

kerem commented

Author

Owner

@Telemaniaka commented on GitHub (Sep 1, 2025):

@atrozgrifo I'm not gonna be of much help :/ I started with MickLesk's script 2 weeks ago.
And I'm already fuzzy on some of the steps I did.

After it finished, Frigate wasn't starting. I saw in /run/shm/logs/frigate/current that I was missing ffmpeg.
Spend some time trying to figure out which version and from where to get it.
Then I found out that it usually gets installed by /opt/frigate/docker/main/install_deps.sh. So I ran that, and manually did the symlink steps from the frigate-install script.
And Frigate finally started.

I moved my frigate config from the old v14 instance with small changes to use CPU detector and no hwaccel for ffmpeg.
And at that point I had a working lxc that I could start and stop to try and figure out why /dev/dri was empty

@Telemaniaka commented on GitHub (Sep 1, 2025): @atrozgrifo I'm not gonna be of much help :/ I started with MickLesk's script 2 weeks ago. And I'm already fuzzy on some of the steps I did. After it finished, Frigate wasn't starting. I saw in `/run/shm/logs/frigate/current` that I was missing `ffmpeg`. Spend some time trying to figure out which version and from where to get it. Then I found out that it usually gets installed by `/opt/frigate/docker/main/install_deps.sh`. So I ran that, and manually did the symlink steps from the frigate-install script. And Frigate finally started. I moved my frigate config from the old v14 instance with small changes to use CPU detector and no hwaccel for ffmpeg. And at that point I had a working lxc that I could start and stop to try and figure out why `/dev/dri` was empty

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 1, 2025):

@Telemaniaka Ok, ya that's different. I was using the current script and although it starts up a LXC it fails later and I'm not sure what's left to do. I can fix the .conf file but that doesn't do anything to recover the install. I guess I could spend a lot of time looking into that but I expect the problem is an easy fix (removing the / that you found) for the good folks creating this.

I have a Coral.ai in another system but I'm trying to do a fresh install on a new box without one, just the CPU.

Thanks

@atrozgrifo commented on GitHub (Sep 1, 2025): @Telemaniaka Ok, ya that's different. I was using the current script and although it starts up a LXC it fails later and I'm not sure what's left to do. I can fix the .conf file but that doesn't do anything to recover the install. I guess I could spend a lot of time looking into that but I expect the problem is an easy fix (removing the / that you found) for the good folks creating this. I have a Coral.ai in another system but I'm trying to do a fresh install on a new box without one, just the CPU. Thanks

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 2, 2025):

@Zixim and @iz8mbw do you get the same error when running the current official script? (Frigate v15):
bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/ct/frigate.sh)"

I'm asking because the error you are getting comes from the iGPU passthrough code that I copied from the existing script. Unfortunately, I can't test myself as I don't have access to a CPU with iGPU (hence why I work on supporting Nvidia GPUs).

@remz1337 commented on GitHub (Sep 2, 2025): @Zixim and @iz8mbw do you get the same error when running the current official script? (Frigate v15): `bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/ct/frigate.sh)"` I'm asking because the error you are getting comes from the iGPU passthrough code that I copied from the existing script. Unfortunately, I can't test myself as I don't have access to a CPU with iGPU (hence why I work on supporting Nvidia GPUs).

kerem commented

Author

Owner

@iz8mbw commented on GitHub (Sep 2, 2025):

@remz1337 with bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/ct/frigate.sh)" it works OK and during the LXC creation there is a question about to enable "VAAPI" passthrough hardware acceleration (I answered Yes to enable it) that I cannot see in your script.
Here the part about "VAAPI":

  Creating a Frigate LXC using the above advanced settings
  💡  Writing configuration to /opt/community-scripts/frigate.conf
  ✔️   Storage local (Free: 26.1GB  Used: 8.4GB) [Template]
  ✔️   Storage local-lvm (Free: 267.3GB  Used: 118.2GB) [Container]
  ✔️   Template debian-11-standard_11.7-1_amd64.tar.zst [local]
  ✔️   LXC Container 108 was successfully created.

 ⚙️  Configuring VAAPI passthrough for LXC container
 ℹ️  VAAPI enables GPU hardware acceleration (e.g., for video transcoding in Jellyfin or Plex).

➤ Automatically mount all available VAAPI devices? [Y/n]: y
  ✔️   Started LXC Container

And in my old Frigate LXC I use the hwaccel_args: preset-vaapi.

@iz8mbw commented on GitHub (Sep 2, 2025): @remz1337 with `bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/ct/frigate.sh)"` it works OK and during the LXC creation there is a question about to enable "VAAPI" passthrough hardware acceleration (I answered Yes to enable it) that I cannot see in your script. Here the part about "VAAPI": ``` Creating a Frigate LXC using the above advanced settings 💡 Writing configuration to /opt/community-scripts/frigate.conf ✔️ Storage local (Free: 26.1GB Used: 8.4GB) [Template] ✔️ Storage local-lvm (Free: 267.3GB Used: 118.2GB) [Container] ✔️ Template debian-11-standard_11.7-1_amd64.tar.zst [local] ✔️ LXC Container 108 was successfully created. ⚙️ Configuring VAAPI passthrough for LXC container ℹ️ VAAPI enables GPU hardware acceleration (e.g., for video transcoding in Jellyfin or Plex). ➤ Automatically mount all available VAAPI devices? [Y/n]: y ✔️ Started LXC Container ``` And in my old Frigate LXC I use the `hwaccel_args: preset-vaapi`.

kerem commented

Author

Owner

@MickLesk commented on GitHub (Sep 2, 2025):

This is also the reason why the script should be in the dev repo and nowhere else. 😄

@MickLesk commented on GitHub (Sep 2, 2025): This is also the reason why the script should be in the dev repo and nowhere else. :smile:

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 2, 2025):

@remz1337 I fully confirm @iz8mbw's answer.
And like you said, the 0.16 install fails because NVIDIA <--> iGPU
Also, the official 0.15 install is saying that TPU (Coral) is being installed, which I can't remember seeing on your 0.16 install.

@Zixim commented on GitHub (Sep 2, 2025): @remz1337 I fully confirm @iz8mbw's answer. And like you said, the 0.16 install fails because NVIDIA <--> iGPU Also, the official 0.15 install is saying that TPU (Coral) is being installed, which I can't remember seeing on your 0.16 install.

kerem commented

PR:
https://github.com/community-scripts/ProxmoxVED/pull/857

Author

Owner

@remz1337 commented on GitHub (Sep 4, 2025):

Thanks for your feedback, that should be enough for me to complete a first PR on the dev repo. You can test it here:
bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/pr-frigate/ct/frigate.sh)"

There is still much improvements to be done in the install script but my first objective is to get it working.

@remz1337 commented on GitHub (Sep 4, 2025): Thanks for your feedback, that should be enough for me to complete a first PR on the dev repo. You can test it here: `bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/pr-frigate/ct/frigate.sh)"` There is still much improvements to be done in the install script but my first objective is to get it working. PR: https://github.com/community-scripts/ProxmoxVED/pull/857

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 4, 2025):

Thanks for your feedback, that should be enough for me to complete a first PR on the dev repo. You can test it here: bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/pr-frigate/ct/frigate.sh)"

It took about 15 minutes and it seemed to run through fully. Congratulations and thank you. Earlier attempts would bail out in a minute.

When I try to open the frigate web page though I just get the busy symbol. I can do a few things like tell it to restart, try to add a camera group, but for the most part it's just that busy symbol. I can't load the configuration editor. That's with the default config.yml.

If I look at the proxmox status it shows the LXC running around 40% of 3 CPUs. If I look at 'top' within the lxc there's very little CPU time used.

The test mp4 file that is used for the default config is the wrong file name. It doesn't exist, there's a different file. Fixing that didn't seem to fix the problem though.
Original: - path: /media/frigate/car-stopping.mp4
Actual: - path: /media/frigate/person-bicycle-car-detection.mp4

I tried using the config.yml.example file and it still just gives the busy symbol.

The /etc/pve/lxc/ *.conf file still has the extra / in the 3 mount entries reported earlier. Is that OK? Normally I don't see that slash. I tried taking it out but it made no difference.

Running on Proxmox 9.06 on an N150 with no extra GPU card, Coral, etc. Browsers tested include Safari on Mac and iPhone and Firefox on Mac.

Edit: I just found this in the frigate log file: /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run: line 20: /usr/lib/ffmpeg/7.0/bin/ffmpeg: No such file or directory

It's right, this doesn't exist, there's no /usr/lib/ffmpeg

@atrozgrifo commented on GitHub (Sep 4, 2025): > Thanks for your feedback, that should be enough for me to complete a first PR on the dev repo. You can test it here: `bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/pr-frigate/ct/frigate.sh)"` It took about 15 minutes and it seemed to run through fully. Congratulations and thank you. Earlier attempts would bail out in a minute. When I try to open the frigate web page though I just get the busy symbol. I can do a few things like tell it to restart, try to add a camera group, but for the most part it's just that busy symbol. I can't load the configuration editor. That's with the default config.yml. If I look at the proxmox status it shows the LXC running around 40% of 3 CPUs. If I look at 'top' within the lxc there's very little CPU time used. The test mp4 file that is used for the default config is the wrong file name. It doesn't exist, there's a different file. Fixing that didn't seem to fix the problem though. Original: - path: /media/frigate/car-stopping.mp4 Actual: - path: /media/frigate/person-bicycle-car-detection.mp4 I tried using the config.yml.example file and it still just gives the busy symbol. The /etc/pve/lxc/ *.conf file still has the extra / in the 3 mount entries reported earlier. Is that OK? Normally I don't see that slash. I tried taking it out but it made no difference. Running on Proxmox 9.06 on an N150 with no extra GPU card, Coral, etc. Browsers tested include Safari on Mac and iPhone and Firefox on Mac. Edit: I just found this in the frigate log file: /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run: line 20: /usr/lib/ffmpeg/7.0/bin/ffmpeg: No such file or directory It's right, this doesn't exist, there's no /usr/lib/ffmpeg

kerem commented

Author

Owner

@rekenaar00 commented on GitHub (Sep 4, 2025):

You can test it here:
bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/pr-frigate/ct/frigate.sh)"

Hi, I am struggeling...I just installed it, but it is not working. Web interface loading left bar and nothing else. Tried to open log, but not working.
Browser dev tools show the following:
i18n-DVPH9xps.js:76 GET http://192.168.50.193:5000/api/config 500 (Internal Server Error)

Then I had to do PCT ENTER since I could not access shell otherwise.
Then I looked in /var/log, but could not identify anything useful, wherecan I look?

@rekenaar00 commented on GitHub (Sep 4, 2025): > You can test it here: > `bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/pr-frigate/ct/frigate.sh)"` Hi, I am struggeling...I just installed it, but it is not working. Web interface loading left bar and nothing else. Tried to open log, but not working. Browser dev tools show the following: i18n-DVPH9xps.js:76 GET http://192.168.50.193:5000/api/config 500 (Internal Server Error) Then I had to do PCT ENTER since I could not access shell otherwise. Then I looked in /var/log, but could not identify anything useful, wherecan I look?

kerem commented

Author

Owner

@iz8mbw commented on GitHub (Sep 4, 2025):

Thanks for your feedback, that should be enough for me to complete a first PR on the dev repo. You can test it here: bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/pr-frigate/ct/frigate.sh)"

There is still much improvements to be done in the install script but my first objective is to get it working.

PR: community-scripts/ProxmoxVED#857

Hello @remz1337.
Just tested and now the installation doesn't provide errors (only some wornings), but when I access to Frigate web interface I receive lot's of HTTP 500 and webpage is broken (always loading).
See here from Chrome console:

@iz8mbw commented on GitHub (Sep 4, 2025): > Thanks for your feedback, that should be enough for me to complete a first PR on the dev repo. You can test it here: `bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/pr-frigate/ct/frigate.sh)"` > > There is still much improvements to be done in the install script but my first objective is to get it working. > > PR: [community-scripts/ProxmoxVED#857](https://github.com/community-scripts/ProxmoxVED/pull/857) Hello @remz1337. Just tested and now the installation doesn't provide errors (only some wornings), but when I access to Frigate web interface I receive lot's of HTTP 500 and webpage is broken (always loading). See here from Chrome console: <img width="540" height="293" alt="Image" src="https://github.com/user-attachments/assets/5e69ca77-65c8-4f6d-aeaf-340851f1c1cf" />

kerem commented

Author

Owner

@Honusnap commented on GitHub (Sep 5, 2025):

Hello all,
Is the install compatible with all version of 0.16.1 ? Rocm, Rk, Std ... etc ?

@Honusnap commented on GitHub (Sep 5, 2025): Hello all, Is the install compatible with all version of 0.16.1 ? Rocm, Rk, Std ... etc ?

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 5, 2025):

for now, the installer errors out at a very early stage :

  🗂️  Enable FUSE Support: no
  🔍  Verbose Mode: yes
  🚀  Creating a Frigate LXC using the above advanced settings
bash: line 267: TEMPLATE_STORAGE: unbound variable
root@pve:~#

@Zixim commented on GitHub (Sep 5, 2025): for now, the installer errors out at a very early stage : ``` 🗂️ Enable FUSE Support: no 🔍 Verbose Mode: yes 🚀 Creating a Frigate LXC using the above advanced settings bash: line 267: TEMPLATE_STORAGE: unbound variable root@pve:~# ```

kerem commented

Author

Owner

@Adrien40 commented on GitHub (Sep 5, 2025):

Hello, which link should I type in the Proxmox shell ? There's remz1337's and MickLesk's too !
Thanks for your work !

@Adrien40 commented on GitHub (Sep 5, 2025): Hello, which link should I type in the Proxmox shell ? There's remz1337's and MickLesk's too ! Thanks for your work !

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 5, 2025):

neither, at this point things are "experimental" and are not expected to function.

@Zixim commented on GitHub (Sep 5, 2025): neither, at this point things are "experimental" and are not expected to function.

kerem commented

Author

Owner

@MickLesk commented on GitHub (Sep 5, 2025):

Hello, which link should I type in the Proxmox shell ? There's remz1337's and MickLesk's too !
Thanks for your work !

The PR from remz in ProxmoxVED is currently in Draft, so i think his is the correct

@MickLesk commented on GitHub (Sep 5, 2025): > Hello, which link should I type in the Proxmox shell ? There's remz1337's and MickLesk's too ! > Thanks for your work ! The PR from remz in ProxmoxVED is currently in Draft, so i think his is the correct

kerem commented

Author

Owner

@mattosaurus commented on GitHub (Sep 8, 2025):

Thanks for your feedback, that should be enough for me to complete a first PR on the dev repo. You can test it here: bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/pr-frigate/ct/frigate.sh)"

There is still much improvements to be done in the install script but my first objective is to get it working.

PR: community-scripts/ProxmoxVED#857

To get this working I needed to manually install ffmpeg, vaapi and the coral tpu usb drivers in the lxc. The draft script didn't work out of the box for me, though that may have been my particualr setup.

apt update
apt install -y ffmpeg

echo "deb https://packages.cloud.google.com/apt coral-edgetpu-stable main" \
    | tee /etc/apt/sources.list.d/coral-edgetpu.list
curl https://packages.cloud.google.com/apt/doc/apt-key.gpg | apt-key add -

apt install -y libedgetpu1-std
apt install -y vainfo intel-media-va-driver i965-va-driver

mkdir -p /usr/lib/ffmpeg/7.0/bin
ln -s /usr/bin/ffmpeg /usr/lib/ffmpeg/7.0/bin/ffmpeg
ln -s /usr/bin/ffprobe /usr/lib/ffmpeg/7.0/bin/ffprobe

@mattosaurus commented on GitHub (Sep 8, 2025): > Thanks for your feedback, that should be enough for me to complete a first PR on the dev repo. You can test it here: `bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/pr-frigate/ct/frigate.sh)"` > > There is still much improvements to be done in the install script but my first objective is to get it working. > > PR: [community-scripts/ProxmoxVED#857](https://github.com/community-scripts/ProxmoxVED/pull/857) To get this working I needed to manually install ffmpeg, vaapi and the coral tpu usb drivers in the lxc. The draft script didn't work out of the box for me, though that may have been my particualr setup. ```bash apt update apt install -y ffmpeg echo "deb https://packages.cloud.google.com/apt coral-edgetpu-stable main" \ | tee /etc/apt/sources.list.d/coral-edgetpu.list curl https://packages.cloud.google.com/apt/doc/apt-key.gpg | apt-key add - apt install -y libedgetpu1-std apt install -y vainfo intel-media-va-driver i965-va-driver mkdir -p /usr/lib/ffmpeg/7.0/bin ln -s /usr/bin/ffmpeg /usr/lib/ffmpeg/7.0/bin/ffmpeg ln -s /usr/bin/ffprobe /usr/lib/ffmpeg/7.0/bin/ffprobe ```

kerem commented

Author

Owner

@iz8mbw commented on GitHub (Sep 9, 2025):

apt install -y libedgetpu1-std
apt install -y vainfo intel-media-va-driver i965-va-driver

still not working for me

@iz8mbw commented on GitHub (Sep 9, 2025): > apt install -y libedgetpu1-std > apt install -y vainfo intel-media-va-driver i965-va-driver still not working for me

kerem commented

Author

Owner

@rommess commented on GitHub (Sep 10, 2025):

Hello,
The script is promising, thank you.
After a infinite loop loading of the web page, i did

apt install -y ffmpeg
apt install -y vainfo intel-media-va-driver i965-va-driver
mkdir -p /usr/lib/ffmpeg/7.0/bin
ln -s /usr/bin/ffmpeg /usr/lib/ffmpeg/7.0/bin/ffmpeg
ln -s /usr/bin/ffprobe /usr/lib/ffmpeg/7.0/bin/ffprobe

And the web page appear.
i'm not using Coral, only GPU.

@rommess commented on GitHub (Sep 10, 2025): Hello, The script is promising, thank you. After a infinite loop loading of the web page, i did ``` apt install -y ffmpeg apt install -y vainfo intel-media-va-driver i965-va-driver mkdir -p /usr/lib/ffmpeg/7.0/bin ln -s /usr/bin/ffmpeg /usr/lib/ffmpeg/7.0/bin/ffmpeg ln -s /usr/bin/ffprobe /usr/lib/ffmpeg/7.0/bin/ffprobe ``` And the web page appear. i'm not using Coral, only GPU.

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 11, 2025):

I have just pushed an update to the script. Can someone try again and let me know if there are still issues?
Thanks

On my end, it works for both TensorRT (Nvidia GPU) and Openvino (CPU), but I can't test with any other hardware (Intel/AMD iGPUs). The script downloads a few additional onnx models, I would appreciate if someone with an iGPU could test and report on performance.

@remz1337 commented on GitHub (Sep 11, 2025): I have just pushed an update to the script. Can someone try again and let me know if there are still issues? Thanks On my end, it works for both TensorRT (Nvidia GPU) and Openvino (CPU), but I can't test with any other hardware (Intel/AMD iGPUs). The script downloads a few additional onnx models, I would appreciate if someone with an iGPU could test and report on performance.

kerem commented

Author

Owner

@iz8mbw commented on GitHub (Sep 11, 2025):

I have just pushed an update to the script. Can someone try again and let me know if there are still issues? Thanks

On my end, it works for both TensorRT (Nvidia GPU) and Openvino (CPU), but I can't test with any other hardware (Intel/AMD iGPUs). The script downloads a few additional onnx models, I would appreciate if someone with an iGPU could test and report on performance.

still HTTP 500 and so infinite loading on the webpage.
Tried also to manually edit /opt/frigate/config/config.yml passing a working configuration but still HTTP 500.

EDIT:
tried also to:

apt install -y ffmpeg
apt install -y vainfo intel-media-va-driver i965-va-driver
mkdir -p /usr/lib/ffmpeg/7.0/bin
ln -s /usr/bin/ffmpeg /usr/lib/ffmpeg/7.0/bin/ffmpeg
ln -s /usr/bin/ffprobe /usr/lib/ffmpeg/7.0/bin/ffprobe

seems better on the web interface but on Frigate log I have:

Unable to read frames from ffmpeg process.
ffmpeg process is not running. exiting capture thread...

@iz8mbw commented on GitHub (Sep 11, 2025): > I have just pushed an update to the script. Can someone try again and let me know if there are still issues? Thanks > > On my end, it works for both TensorRT (Nvidia GPU) and Openvino (CPU), but I can't test with any other hardware (Intel/AMD iGPUs). The script downloads a few additional onnx models, I would appreciate if someone with an iGPU could test and report on performance. still HTTP 500 and so infinite loading on the webpage. Tried also to manually edit `/opt/frigate/config/config.yml` passing a working configuration but still HTTP 500. EDIT: tried also to: ``` apt install -y ffmpeg apt install -y vainfo intel-media-va-driver i965-va-driver mkdir -p /usr/lib/ffmpeg/7.0/bin ln -s /usr/bin/ffmpeg /usr/lib/ffmpeg/7.0/bin/ffmpeg ln -s /usr/bin/ffprobe /usr/lib/ffmpeg/7.0/bin/ffprobe ``` seems better on the web interface but on Frigate log I have: ``` Unable to read frames from ffmpeg process. ffmpeg process is not running. exiting capture thread... ```

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 11, 2025):

@remz1337 Here's the output of the installation on an iGPU host, sadly still failing for me :

  ⚙️  Using Default Settings on node pve (🔍  Verbose)
  🆔  Container ID: 100
  🖥️  Operating System: debian (12)
  📦  Container Type: Privileged
  💾  Disk Size: 20 GB
  🧠  CPU Cores: 4
  🛠️  RAM Size: 4096 MiB
  🚀  Creating a Frigate LXC using the above default settings
  
  ✔️   Storage iso (Free: 842.1GB  Used: 5.9GB) [Template]
  ✔️   Storage lxc-zfs (Free: 538.0GB  Used: 20.4GB) [Container]
  ✔️   Storage 'lxc-zfs' supports 'rootdir'
  ✔️   Template storage 'iso' supports 'vztmpl'
  ✔️   Template debian-12-standard_12.7-1_amd64.tar.zst [local]
  ✔️   Template debian-12-standard_12.7-1_amd64.tar.zst is present and valid.
   💡   Local template is outdated: debian-12-standard_12.7-1_amd64.tar.zst (latest available: debian-12-standard_12.12-1_amd64.tar.zst)
  ✔️   Template download successful.
  ✔️   Template debian-12-standard_12.12-1_amd64.tar.zst is ready for container creation.
  ✔️   LXC Container 100 was successfully created.
  ✔️   Started LXC Container
  ✔️   Network in LXC is reachable (ping)
  ⏳   Customizing LXC Container
  ✔️   Set up Container OS
  ✔️   Network Connected: 192.168.11.198 
  ✔️   IPv4 Internet Connected
   ✖️   IPv6 Internet Not Connected
  ✔️   Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) git.community-scripts.org:(✔️ )
Hit:1 http://security.debian.org bookworm-security InRelease

... output deleted by user to shorten the posting on GH ...

Setting up intel-opencl-icd (22.43.24595.41-1) ...
Processing triggers for libc-bin (2.36-9+deb12u13) ...
Processing triggers for man-db (2.11.2-2) ...
chgrp: cannot access '/dev/dri': No such file or directory

[ERROR] in line 24: exit code 0: while executing command chgrp video /dev/dri

[ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)"

/dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required
root@pve:~#

When the install script prompted me regarding pass-through, I selected the iGPU and not the fb0 device.

root@frigate:~# ls -l /dev/dri
ls: cannot access '/dev/dri': No such file or directory
root@frigate:~#

Yet the LXC conf file has the mount :
lxc.mount.entry: /dev/dri/renderD128 /dev/dri/renderD128 none bind,optional,create=file

Is the mount added to the config after the container is started ? That would cause /dev/dri to be missing.

Perhaps the Jellyfin install can help, it install fine on iGPU hosts.

@Zixim commented on GitHub (Sep 11, 2025): @remz1337 Here's the output of the installation on an iGPU host, sadly still failing for me : ``` ⚙️ Using Default Settings on node pve (🔍 Verbose) 🆔 Container ID: 100 🖥️ Operating System: debian (12) 📦 Container Type: Privileged 💾 Disk Size: 20 GB 🧠 CPU Cores: 4 🛠️ RAM Size: 4096 MiB 🚀 Creating a Frigate LXC using the above default settings ✔️ Storage iso (Free: 842.1GB Used: 5.9GB) [Template] ✔️ Storage lxc-zfs (Free: 538.0GB Used: 20.4GB) [Container] ✔️ Storage 'lxc-zfs' supports 'rootdir' ✔️ Template storage 'iso' supports 'vztmpl' ✔️ Template debian-12-standard_12.7-1_amd64.tar.zst [local] ✔️ Template debian-12-standard_12.7-1_amd64.tar.zst is present and valid. 💡 Local template is outdated: debian-12-standard_12.7-1_amd64.tar.zst (latest available: debian-12-standard_12.12-1_amd64.tar.zst) ✔️ Template download successful. ✔️ Template debian-12-standard_12.12-1_amd64.tar.zst is ready for container creation. ✔️ LXC Container 100 was successfully created. ✔️ Started LXC Container ✔️ Network in LXC is reachable (ping) ⏳ Customizing LXC Container ✔️ Set up Container OS ✔️ Network Connected: 192.168.11.198 ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) git.community-scripts.org:(✔️ ) Hit:1 http://security.debian.org bookworm-security InRelease ... output deleted by user to shorten the posting on GH ... Setting up intel-opencl-icd (22.43.24595.41-1) ... Processing triggers for libc-bin (2.36-9+deb12u13) ... Processing triggers for man-db (2.11.2-2) ... chgrp: cannot access '/dev/dri': No such file or directory [ERROR] in line 24: exit code 0: while executing command chgrp video /dev/dri [ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)" /dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required root@pve:~# ```` When the install script prompted me regarding pass-through, I selected the iGPU and not the fb0 device. ``` root@frigate:~# ls -l /dev/dri ls: cannot access '/dev/dri': No such file or directory root@frigate:~# ``` Yet the LXC conf file has the mount : `lxc.mount.entry: /dev/dri/renderD128 /dev/dri/renderD128 none bind,optional,create=file` Is the mount added to the config after the container is started ? That would cause /dev/dri to be missing. Perhaps the Jellyfin install can help, it install fine on iGPU hosts.

kerem commented

Author

Owner

@Adrien40 commented on GitHub (Sep 11, 2025):

Hello, I just tested it, and it seems the installation went well !
Here's my setup : Intel Twin Lake N150 sous Proxmox 9.0.6
NiPoGi Essenx E1 Mini PC W-11 PRO Intel Twin Lake N150(Refreshed N100)16Go RAM 512Go M.2 SSD, Ordinateur de Bureau 4K@60HZ UHD Dual Display HDMI+DP 1.4/Dual Bande WiFi/BT4.2/RJ45/Tour PC VESA Supporté

On the other hand, the Frigate documentation mentions:

ports: - "8971:8971" # - "5000:5000" # Internal unauthenticated access. Expose carefully. - "8554:8554" # RTSP feeds - "8555:8555/tcp" # WebRTC over tcp - "8555:8555/udp" # WebRTC over udp

so why use port 5000 ?

`
⚙️ Using Default Settings on node proxmox
🆔 Container ID: 103
🖥️ Operating System: debian (12)
📦 Container Type: Privileged
💾 Disk Size: 20 GB
🧠 CPU Cores: 4
🛠️ RAM Size: 4096 MiB
🚀 Creating a Frigate LXC using the above default settings

✔️ Storage local (Free: 53.4GB Used: 33.7GB) [Template]
✔️ Storage local-lvm (Free: 305.9GB Used: 34.0GB) [Container]
✔️ Storage 'local-lvm' supports 'rootdir'
✔️ Template storage 'local' supports 'vztmpl'
✔️ Template debian-12-standard_12.12-1_amd64.tar.zst [local]
✔️ Template debian-12-standard_12.12-1_amd64.tar.zst is present and valid.
✔️ Template debian-12-standard_12.12-1_amd64.tar.zst is ready for container creation.
✔️ LXC Container 103 was successfully created.
✔️ Started LXC Container
✔️ Network in LXC is reachable (ping)
✔️ Customized LXC Container
✔️ Set up Container OS
✔️ Network Connected: 192.168.1.9 2a01:e0a:c36:e660:be24:11ff:fe8b:6ff1
✔️ IPv4 Internet Connected
✔️ IPv6 Internet Connected
✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) git.community-scripts.org:(✔️ )
✔️ Updated Container OS
✔️ Installed Dependencies
✔️ Setup Python3
✔️ Setup Node.js 22
✔️ Installed Node.js modules: yarn
✔️ Deployed: go2rtc (1.9.9)
✔️ Deployed: frigate (0.16.1)
✔️ Deployed: libusb (1.0.29)
✔️ Python venv ready
✔️ Web UI built
✔️ Config ready
✔️ AVX Support Detected
✔️ Installed Openvino Object Detection Model
⠇ Building and Installing libUSB without udevautoreconf: export WARNINGS=
autoreconf: Entering directory '.'
autoreconf: configure.ac: not using Gettext
⠇ Building and Installing libUSB without udevautoreconf: running: aclocal --force -I m4
⠇ Building and Installing libUSB without udevautoreconf: configure.ac: tracing
⠙ Building and Installing libUSB without udevautoreconf: running: libtoolize --copy --force
libtoolize: putting auxiliary files in '.'.
libtoolize: copying file './ltmain.sh'
⠹ Building and Installing libUSB without udevlibtoolize: putting macros in AC_CONFIG_MACRO_DIRS, 'm4'.
libtoolize: copying file 'm4/libtool.m4'
libtoolize: copying file 'm4/ltoptions.m4'
libtoolize: copying file 'm4/ltsugar.m4'
libtoolize: copying file 'm4/ltversion.m4'
libtoolize: copying file 'm4/lt~obsolete.m4'
autoreconf: configure.ac: not using Intltool
autoreconf: configure.ac: not using Gtkdoc
autoreconf: running: aclocal --force -I m4
⠋ Building and Installing libUSB without udevautoreconf: running: /usr/bin/autoconf --force
⠴ Building and Installing libUSB without udevautoreconf: running: /usr/bin/autoheader --force
⠏ Building and Installing libUSB without udevautoreconf: running: automake --add-missing --copy --force-missing
⠙ Building and Installing libUSB without udevconfigure.ac:23: installing './compile'
configure.ac:27: installing './config.guess'
configure.ac:27: installing './config.sub'
configure.ac:26: installing './install-sh'
configure.ac:26: installing './missing'
Makefile.am: installing './INSTALL'
examples/Makefile.am: installing './depcomp'
⠹ Building and Installing libUSB without udevparallel-tests: installing './test-driver'
⠸ Building and Installing libUSB without udevautoreconf: Leaving directory '.'
checking for gcc... gcc
⠼ Building and Installing libUSB without udevchecking whether the C compiler works... yes
checking for C compiler default output file name... a.out
checking for suffix of executables...
⠴ Building and Installing libUSB without udevno
checking for suffix of object files... o
checking whether the compiler supports GNU C... yes
checking whether gcc accepts -g... yes
⠦ Building and Installing libUSB without udevnone needed
checking whether gcc understands -c and -o together... yes
checking for g++... g++
⠇ Building and Installing libUSB without udevyes
checking whether g++ accepts -g... yes
⠋ Building and Installing libUSB without udevnone needed
checking for inline... inline
⠙ Building and Installing libUSB without udev/usr/bin/install -c
checking whether build environment is sane... yes
checking for a race-free mkdir -p... /usr/bin/mkdir -p
checking for gawk... no
checking for mawk... mawk
checking whether make sets $(MAKE)... yes
checking whether make supports the include directive... yes (GNU style)
checking whether make supports nested variables... yes
⠹ Building and Installing libUSB without udevgcc3
checking dependency style of g++... gcc3
⠸ Building and Installing libUSB without udevx86_64-pc-linux-gnu
checking host system type... x86_64-pc-linux-gnu
checking how to print strings... printf
checking for a sed that does not truncate output... /usr/bin/sed
checking for grep that handles long lines and -e... /usr/bin/grep
checking for egrep... /usr/bin/grep -E
checking for fgrep... /usr/bin/grep -F
checking for ld used by gcc... /usr/bin/ld
checking if the linker (/usr/bin/ld) is GNU ld... yes
checking for BSD- or MS-compatible name lister (nm)... /usr/bin/nm -B
checking the name lister (/usr/bin/nm -B) interface... BSD nm
checking whether ln -s works... yes
checking the maximum length of command line arguments... 1572864
checking how to convert x86_64-pc-linux-gnu file names to x86_64-pc-linux-gnu format... func_convert_file_noop
checking how to convert x86_64-pc-linux-gnu file names to toolchain format... func_convert_file_noop
checking for /usr/bin/ld option to reload object files... -r
checking for file... file
checking for objdump... objdump
checking how to recognize dependent libraries... pass_all
checking for dlltool... no
checking how to associate runtime and link libraries... printf %s\n
checking for ar... ar
⠼ Building and Installing libUSB without udev@
checking for strip... strip
checking for ranlib... ranlib
checking command to parse /usr/bin/nm -B output from gcc object... ok
checking for sysroot... no
checking for a working dd... /usr/bin/dd
checking how to truncate binary pipes... /usr/bin/dd bs=4096 count=1
⠴ Building and Installing libUSB without udevchecking for mt... mt
checking if mt is a manifest tool... no
checking for stdio.h... yes
checking for stdlib.h... yes
checking for string.h... yes
⠦ Building and Installing libUSB without udevyes
checking for stdint.h... yes
checking for strings.h... yes
⠧ Building and Installing libUSB without udevyes
checking for sys/types.h... yes
checking for unistd.h... yes
⠇ Building and Installing libUSB without udevyes
checking for objdir... .libs
checking if gcc supports -fno-rtti -fno-exceptions... no
checking for gcc option to produce PIC... -fPIC -DPIC
checking if gcc PIC flag -fPIC -DPIC works... yes
⠏ Building and Installing libUSB without udevyes
⠋ Building and Installing libUSB without udevyes
checking if gcc supports -c -o file.o... (cached) yes
checking whether the gcc linker (/usr/bin/ld -m elf_x86_64) supports shared libraries... yes
checking whether -lc should be explicitly linked in... no
checking dynamic linker characteristics... GNU/Linux ld.so
checking how to hardcode library paths into programs... immediate
checking whether stripping libraries is possible... yes
checking if libtool supports shared libraries... yes
checking whether to build shared libraries... yes
checking whether to build static libraries... yes
⠙ Building and Installing libUSB without udevg++ -E
⠹ Building and Installing libUSB without udevchecking for ld used by g++... /usr/bin/ld -m elf_x86_64
checking if the linker (/usr/bin/ld -m elf_x86_64) is GNU ld... yes
checking whether the g++ linker (/usr/bin/ld -m elf_x86_64) supports shared libraries... yes
checking for g++ option to produce PIC... -fPIC -DPIC
checking if g++ PIC flag -fPIC -DPIC works... yes
⠸ Building and Installing libUSB without udevyes
checking if g++ supports -c -o file.o... yes
checking if g++ supports -c -o file.o... (cached) yes
checking whether the g++ linker (/usr/bin/ld -m elf_x86_64) supports shared libraries... yes
checking dynamic linker characteristics... (cached) GNU/Linux ld.so
checking how to hardcode library paths into programs... immediate
checking for windres... no
⠼ Building and Installing libUSB without udevchecking whether make supports nested variables... (cached) yes
checking if gcc supports -std=gnu11... yes
checking operating system... Linux
checking for nfds_t... yes
⠴ Building and Installing libUSB without udevyes
checking if gcc recognizes -pthread... yes
⠦ Building and Installing libUSB without udevnone required
checking whether __atomic_load_n is supported... yes
⠧ Building and Installing libUSB without udevnone required
checking for pthread_setname_np... yes
⠇ Building and Installing libUSB without udevyes
checking for linux/netlink.h... yes
checking for sys/socket.h... yes
⠏ Building and Installing libUSB without udevyes
checking for clock_gettime... yes
⠋ Building and Installing libUSB without udevnone needed
checking whether CLOCK_MONOTONIC is declared... yes
checking for pthread_condattr_setclock... yes
⠙ Building and Installing libUSB without udevyes
checking whether EFD_NONBLOCK is declared... yes
checking whether EFD_CLOEXEC is declared... yes
checking for eventfd... yes
checking whether to use eventfd for signalling... yes
⠹ Building and Installing libUSB without udevyes
checking whether TFD_NONBLOCK is declared... yes
checking whether TFD_CLOEXEC is declared... yes
checking for timerfd_create... yes
checking whether to use timerfd for timing... yes
⠸ Building and Installing libUSB without udevchecking that generated files are newer than configure... done
configure: creating ./config.status
⠇ Building and Installing libUSB without udevconfig.status: creating libusb-1.0.pc
config.status: creating Makefile
config.status: creating libusb/Makefile
config.status: creating examples/Makefile
⠏ Building and Installing libUSB without udevconfig.status: creating tests/Makefile
config.status: creating doc/Makefile
config.status: creating doc/doxygen.cfg
config.status: creating config.h
config.status: executing depfiles commands
⠙ Building and Installing libUSB without udevconfig.status: executing libtool commands
make all-recursive
make[1]: Entering directory '/tmp/libusb-1.0.29'
⠹ Building and Installing libUSB without udevMaking all in libusb
make[2]: Entering directory '/tmp/libusb-1.0.29/libusb'
CC core.lo
CC descriptor.lo
CC hotplug.lo
CC io.lo
⠏ Building and Installing libUSB without udev CC strerror.lo
⠹ Building and Installing libUSB without udev CC sync.lo
⠴ Building and Installing libUSB without udev CC os/events_posix.lo
⠦ Building and Installing libUSB without udev CC os/threads_posix.lo
⠧ Building and Installing libUSB without udev CC os/linux_usbfs.lo
⠇ Building and Installing libUSB without udev CC os/linux_netlink.lo
⠴ Building and Installing libUSB without udev CCLD libusb-1.0.la
⠇ Building and Installing libUSB without udevmake[2]: Leaving directory '/tmp/libusb-1.0.29/libusb'
make[2]: Entering directory '/tmp/libusb-1.0.29'
make[2]: Leaving directory '/tmp/libusb-1.0.29'
make[1]: Leaving directory '/tmp/libusb-1.0.29'
Making install in libusb
make[1]: Entering directory '/tmp/libusb-1.0.29/libusb'
make[2]: Entering directory '/tmp/libusb-1.0.29/libusb'
/usr/bin/mkdir -p '/usr/local/lib'
/bin/bash ../libtool --mode=install /usr/bin/install -c libusb-1.0.la '/usr/local/lib'
⠏ Building and Installing libUSB without udevlibtool: install: /usr/bin/install -c .libs/libusb-1.0.so.0.5.0 /usr/local/lib/libusb-1.0.so.0.5.0
libtool: install: (cd /usr/local/lib && { ln -s -f libusb-1.0.so.0.5.0 libusb-1.0.so.0 || { rm -f libusb-1.0.so.0 && ln -s libusb-1.0.so.0.5.0 libusb-1.0.so.0; }; })
libtool: install: (cd /usr/local/lib && { ln -s -f libusb-1.0.so.0.5.0 libusb-1.0.so || { rm -f libusb-1.0.so && ln -s libusb-1.0.so.0.5.0 libusb-1.0.so; }; })
libtool: install: /usr/bin/install -c .libs/libusb-1.0.lai /usr/local/lib/libusb-1.0.la
libtool: install: /usr/bin/install -c .libs/libusb-1.0.a /usr/local/lib/libusb-1.0.a
libtool: install: chmod 644 /usr/local/lib/libusb-1.0.a
libtool: install: ranlib /usr/local/lib/libusb-1.0.a
libtool: finish: PATH="/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/sbin" ldconfig -n /usr/local/lib

Libraries have been installed in:
/usr/local/lib

If you ever happen to want to link against installed libraries
in a given directory, LIBDIR, you must either use libtool, and
specify the full pathname of the library, or use the '-LLIBDIR'
flag during linking and do at least one of the following:

add LIBDIR to the 'LD_LIBRARY_PATH' environment variable
during execution
add LIBDIR to the 'LD_RUN_PATH' environment variable
during linking
use the '-Wl,-rpath -Wl,LIBDIR' linker flag
have your system administrator add LIBDIR to '/etc/ld.so.conf'

See any operating system documentation about shared libraries for
more information, such as the ld(1) and ld.so(8) manual pages.

/usr/bin/mkdir -p '/usr/local/include/libusb-1.0'
/usr/bin/install -c -m 644 libusb.h '/usr/local/include/libusb-1.0'
make[2]: Leaving directory '/tmp/libusb-1.0.29/libusb'
make[1]: Leaving directory '/tmp/libusb-1.0.29/libusb'
make[1]: Entering directory '/tmp/libusb-1.0.29'
make[2]: Entering directory '/tmp/libusb-1.0.29'
make[2]: Nothing to be done for 'install-exec-am'.
⠋ Building and Installing libUSB without udev /usr/bin/mkdir -p '/usr/local/lib/pkgconfig'
/usr/bin/install -c -m 644 libusb-1.0.pc '/usr/local/lib/pkgconfig'
make[2]: Leaving directory '/tmp/libusb-1.0.29'
make[1]: Leaving directory '/tmp/libusb-1.0.29'
✔️ Installed libUSB without udev
✔️ Installed Coral Object Detection Model
✔️ Installed Tempio
✔️ Built Nginx
✔️ Configured Services
✔️ Completed Successfully!

🚀 Frigate setup has been successfully initialized!
💡 Access it using the following URL:
🌐 http://192.168.1.9:5000`

@Adrien40 commented on GitHub (Sep 11, 2025): Hello, I just tested it, and it seems the installation went well ! Here's my setup : Intel Twin Lake N150 sous Proxmox 9.0.6 `NiPoGi Essenx E1 Mini PC W-11 PRO Intel Twin Lake N150(Refreshed N100)16Go RAM 512Go M.2 SSD, Ordinateur de Bureau 4K@60HZ UHD Dual Display HDMI+DP 1.4/Dual Bande WiFi/BT4.2/RJ45/Tour PC VESA Supporté` On the other hand, the Frigate documentation mentions: ` ports: - "8971:8971" # - "5000:5000" # Internal unauthenticated access. Expose carefully. - "8554:8554" # RTSP feeds - "8555:8555/tcp" # WebRTC over tcp - "8555:8555/udp" # WebRTC over udp` so why use port 5000 ? ` ⚙️ Using Default Settings on node proxmox 🆔 Container ID: 103 🖥️ Operating System: debian (12) 📦 Container Type: Privileged 💾 Disk Size: 20 GB 🧠 CPU Cores: 4 🛠️ RAM Size: 4096 MiB 🚀 Creating a Frigate LXC using the above default settings ✔️ Storage local (Free: 53.4GB Used: 33.7GB) [Template] ✔️ Storage local-lvm (Free: 305.9GB Used: 34.0GB) [Container] ✔️ Storage 'local-lvm' supports 'rootdir' ✔️ Template storage 'local' supports 'vztmpl' ✔️ Template debian-12-standard_12.12-1_amd64.tar.zst [local] ✔️ Template debian-12-standard_12.12-1_amd64.tar.zst is present and valid. ✔️ Template debian-12-standard_12.12-1_amd64.tar.zst is ready for container creation. ✔️ LXC Container 103 was successfully created. ✔️ Started LXC Container ✔️ Network in LXC is reachable (ping) ✔️ Customized LXC Container ✔️ Set up Container OS ✔️ Network Connected: 192.168.1.9 2a01:e0a:c36:e660:be24:11ff:fe8b:6ff1 ✔️ IPv4 Internet Connected ✔️ IPv6 Internet Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) git.community-scripts.org:(✔️ ) ✔️ Updated Container OS ✔️ Installed Dependencies ✔️ Setup Python3 ✔️ Setup Node.js 22 ✔️ Installed Node.js modules: yarn ✔️ Deployed: go2rtc (1.9.9) ✔️ Deployed: frigate (0.16.1) ✔️ Deployed: libusb (1.0.29) ✔️ Python venv ready ✔️ Web UI built ✔️ Config ready ✔️ AVX Support Detected ✔️ Installed Openvino Object Detection Model ⠇ Building and Installing libUSB without udevautoreconf: export WARNINGS= autoreconf: Entering directory '.' autoreconf: configure.ac: not using Gettext ⠇ Building and Installing libUSB without udevautoreconf: running: aclocal --force -I m4 ⠇ Building and Installing libUSB without udevautoreconf: configure.ac: tracing ⠙ Building and Installing libUSB without udevautoreconf: running: libtoolize --copy --force libtoolize: putting auxiliary files in '.'. libtoolize: copying file './ltmain.sh' ⠹ Building and Installing libUSB without udevlibtoolize: putting macros in AC_CONFIG_MACRO_DIRS, 'm4'. libtoolize: copying file 'm4/libtool.m4' libtoolize: copying file 'm4/ltoptions.m4' libtoolize: copying file 'm4/ltsugar.m4' libtoolize: copying file 'm4/ltversion.m4' libtoolize: copying file 'm4/lt~obsolete.m4' autoreconf: configure.ac: not using Intltool autoreconf: configure.ac: not using Gtkdoc autoreconf: running: aclocal --force -I m4 ⠋ Building and Installing libUSB without udevautoreconf: running: /usr/bin/autoconf --force ⠴ Building and Installing libUSB without udevautoreconf: running: /usr/bin/autoheader --force ⠏ Building and Installing libUSB without udevautoreconf: running: automake --add-missing --copy --force-missing ⠙ Building and Installing libUSB without udevconfigure.ac:23: installing './compile' configure.ac:27: installing './config.guess' configure.ac:27: installing './config.sub' configure.ac:26: installing './install-sh' configure.ac:26: installing './missing' Makefile.am: installing './INSTALL' examples/Makefile.am: installing './depcomp' ⠹ Building and Installing libUSB without udevparallel-tests: installing './test-driver' ⠸ Building and Installing libUSB without udevautoreconf: Leaving directory '.' checking for gcc... gcc ⠼ Building and Installing libUSB without udevchecking whether the C compiler works... yes checking for C compiler default output file name... a.out checking for suffix of executables... ⠴ Building and Installing libUSB without udevno checking for suffix of object files... o checking whether the compiler supports GNU C... yes checking whether gcc accepts -g... yes ⠦ Building and Installing libUSB without udevnone needed checking whether gcc understands -c and -o together... yes checking for g++... g++ ⠇ Building and Installing libUSB without udevyes checking whether g++ accepts -g... yes ⠋ Building and Installing libUSB without udevnone needed checking for inline... inline ⠙ Building and Installing libUSB without udev/usr/bin/install -c checking whether build environment is sane... yes checking for a race-free mkdir -p... /usr/bin/mkdir -p checking for gawk... no checking for mawk... mawk checking whether make sets $(MAKE)... yes checking whether make supports the include directive... yes (GNU style) checking whether make supports nested variables... yes ⠹ Building and Installing libUSB without udevgcc3 checking dependency style of g++... gcc3 ⠸ Building and Installing libUSB without udevx86_64-pc-linux-gnu checking host system type... x86_64-pc-linux-gnu checking how to print strings... printf checking for a sed that does not truncate output... /usr/bin/sed checking for grep that handles long lines and -e... /usr/bin/grep checking for egrep... /usr/bin/grep -E checking for fgrep... /usr/bin/grep -F checking for ld used by gcc... /usr/bin/ld checking if the linker (/usr/bin/ld) is GNU ld... yes checking for BSD- or MS-compatible name lister (nm)... /usr/bin/nm -B checking the name lister (/usr/bin/nm -B) interface... BSD nm checking whether ln -s works... yes checking the maximum length of command line arguments... 1572864 checking how to convert x86_64-pc-linux-gnu file names to x86_64-pc-linux-gnu format... func_convert_file_noop checking how to convert x86_64-pc-linux-gnu file names to toolchain format... func_convert_file_noop checking for /usr/bin/ld option to reload object files... -r checking for file... file checking for objdump... objdump checking how to recognize dependent libraries... pass_all checking for dlltool... no checking how to associate runtime and link libraries... printf %s\n checking for ar... ar ⠼ Building and Installing libUSB without udev@ checking for strip... strip checking for ranlib... ranlib checking command to parse /usr/bin/nm -B output from gcc object... ok checking for sysroot... no checking for a working dd... /usr/bin/dd checking how to truncate binary pipes... /usr/bin/dd bs=4096 count=1 ⠴ Building and Installing libUSB without udevchecking for mt... mt checking if mt is a manifest tool... no checking for stdio.h... yes checking for stdlib.h... yes checking for string.h... yes ⠦ Building and Installing libUSB without udevyes checking for stdint.h... yes checking for strings.h... yes ⠧ Building and Installing libUSB without udevyes checking for sys/types.h... yes checking for unistd.h... yes ⠇ Building and Installing libUSB without udevyes checking for objdir... .libs checking if gcc supports -fno-rtti -fno-exceptions... no checking for gcc option to produce PIC... -fPIC -DPIC checking if gcc PIC flag -fPIC -DPIC works... yes ⠏ Building and Installing libUSB without udevyes ⠋ Building and Installing libUSB without udevyes checking if gcc supports -c -o file.o... (cached) yes checking whether the gcc linker (/usr/bin/ld -m elf_x86_64) supports shared libraries... yes checking whether -lc should be explicitly linked in... no checking dynamic linker characteristics... GNU/Linux ld.so checking how to hardcode library paths into programs... immediate checking whether stripping libraries is possible... yes checking if libtool supports shared libraries... yes checking whether to build shared libraries... yes checking whether to build static libraries... yes ⠙ Building and Installing libUSB without udevg++ -E ⠹ Building and Installing libUSB without udevchecking for ld used by g++... /usr/bin/ld -m elf_x86_64 checking if the linker (/usr/bin/ld -m elf_x86_64) is GNU ld... yes checking whether the g++ linker (/usr/bin/ld -m elf_x86_64) supports shared libraries... yes checking for g++ option to produce PIC... -fPIC -DPIC checking if g++ PIC flag -fPIC -DPIC works... yes ⠸ Building and Installing libUSB without udevyes checking if g++ supports -c -o file.o... yes checking if g++ supports -c -o file.o... (cached) yes checking whether the g++ linker (/usr/bin/ld -m elf_x86_64) supports shared libraries... yes checking dynamic linker characteristics... (cached) GNU/Linux ld.so checking how to hardcode library paths into programs... immediate checking for windres... no ⠼ Building and Installing libUSB without udevchecking whether make supports nested variables... (cached) yes checking if gcc supports -std=gnu11... yes checking operating system... Linux checking for nfds_t... yes ⠴ Building and Installing libUSB without udevyes checking if gcc recognizes -pthread... yes ⠦ Building and Installing libUSB without udevnone required checking whether __atomic_load_n is supported... yes ⠧ Building and Installing libUSB without udevnone required checking for pthread_setname_np... yes ⠇ Building and Installing libUSB without udevyes checking for linux/netlink.h... yes checking for sys/socket.h... yes ⠏ Building and Installing libUSB without udevyes checking for clock_gettime... yes ⠋ Building and Installing libUSB without udevnone needed checking whether CLOCK_MONOTONIC is declared... yes checking for pthread_condattr_setclock... yes ⠙ Building and Installing libUSB without udevyes checking whether EFD_NONBLOCK is declared... yes checking whether EFD_CLOEXEC is declared... yes checking for eventfd... yes checking whether to use eventfd for signalling... yes ⠹ Building and Installing libUSB without udevyes checking whether TFD_NONBLOCK is declared... yes checking whether TFD_CLOEXEC is declared... yes checking for timerfd_create... yes checking whether to use timerfd for timing... yes ⠸ Building and Installing libUSB without udevchecking that generated files are newer than configure... done configure: creating ./config.status ⠇ Building and Installing libUSB without udevconfig.status: creating libusb-1.0.pc config.status: creating Makefile config.status: creating libusb/Makefile config.status: creating examples/Makefile ⠏ Building and Installing libUSB without udevconfig.status: creating tests/Makefile config.status: creating doc/Makefile config.status: creating doc/doxygen.cfg config.status: creating config.h config.status: executing depfiles commands ⠙ Building and Installing libUSB without udevconfig.status: executing libtool commands make all-recursive make[1]: Entering directory '/tmp/libusb-1.0.29' ⠹ Building and Installing libUSB without udevMaking all in libusb make[2]: Entering directory '/tmp/libusb-1.0.29/libusb' CC core.lo CC descriptor.lo CC hotplug.lo CC io.lo ⠏ Building and Installing libUSB without udev CC strerror.lo ⠹ Building and Installing libUSB without udev CC sync.lo ⠴ Building and Installing libUSB without udev CC os/events_posix.lo ⠦ Building and Installing libUSB without udev CC os/threads_posix.lo ⠧ Building and Installing libUSB without udev CC os/linux_usbfs.lo ⠇ Building and Installing libUSB without udev CC os/linux_netlink.lo ⠴ Building and Installing libUSB without udev CCLD libusb-1.0.la ⠇ Building and Installing libUSB without udevmake[2]: Leaving directory '/tmp/libusb-1.0.29/libusb' make[2]: Entering directory '/tmp/libusb-1.0.29' make[2]: Leaving directory '/tmp/libusb-1.0.29' make[1]: Leaving directory '/tmp/libusb-1.0.29' Making install in libusb make[1]: Entering directory '/tmp/libusb-1.0.29/libusb' make[2]: Entering directory '/tmp/libusb-1.0.29/libusb' /usr/bin/mkdir -p '/usr/local/lib' /bin/bash ../libtool --mode=install /usr/bin/install -c libusb-1.0.la '/usr/local/lib' ⠏ Building and Installing libUSB without udevlibtool: install: /usr/bin/install -c .libs/libusb-1.0.so.0.5.0 /usr/local/lib/libusb-1.0.so.0.5.0 libtool: install: (cd /usr/local/lib && { ln -s -f libusb-1.0.so.0.5.0 libusb-1.0.so.0 || { rm -f libusb-1.0.so.0 && ln -s libusb-1.0.so.0.5.0 libusb-1.0.so.0; }; }) libtool: install: (cd /usr/local/lib && { ln -s -f libusb-1.0.so.0.5.0 libusb-1.0.so || { rm -f libusb-1.0.so && ln -s libusb-1.0.so.0.5.0 libusb-1.0.so; }; }) libtool: install: /usr/bin/install -c .libs/libusb-1.0.lai /usr/local/lib/libusb-1.0.la libtool: install: /usr/bin/install -c .libs/libusb-1.0.a /usr/local/lib/libusb-1.0.a libtool: install: chmod 644 /usr/local/lib/libusb-1.0.a libtool: install: ranlib /usr/local/lib/libusb-1.0.a libtool: finish: PATH="/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/sbin" ldconfig -n /usr/local/lib ---------------------------------------------------------------------- Libraries have been installed in: /usr/local/lib If you ever happen to want to link against installed libraries in a given directory, LIBDIR, you must either use libtool, and specify the full pathname of the library, or use the '-LLIBDIR' flag during linking and do at least one of the following: - add LIBDIR to the 'LD_LIBRARY_PATH' environment variable during execution - add LIBDIR to the 'LD_RUN_PATH' environment variable during linking - use the '-Wl,-rpath -Wl,LIBDIR' linker flag - have your system administrator add LIBDIR to '/etc/ld.so.conf' See any operating system documentation about shared libraries for more information, such as the ld(1) and ld.so(8) manual pages. ---------------------------------------------------------------------- /usr/bin/mkdir -p '/usr/local/include/libusb-1.0' /usr/bin/install -c -m 644 libusb.h '/usr/local/include/libusb-1.0' make[2]: Leaving directory '/tmp/libusb-1.0.29/libusb' make[1]: Leaving directory '/tmp/libusb-1.0.29/libusb' make[1]: Entering directory '/tmp/libusb-1.0.29' make[2]: Entering directory '/tmp/libusb-1.0.29' make[2]: Nothing to be done for 'install-exec-am'. ⠋ Building and Installing libUSB without udev /usr/bin/mkdir -p '/usr/local/lib/pkgconfig' /usr/bin/install -c -m 644 libusb-1.0.pc '/usr/local/lib/pkgconfig' make[2]: Leaving directory '/tmp/libusb-1.0.29' make[1]: Leaving directory '/tmp/libusb-1.0.29' ✔️ Installed libUSB without udev ✔️ Installed Coral Object Detection Model ✔️ Installed Tempio ✔️ Built Nginx ✔️ Configured Services ✔️ Completed Successfully! 🚀 Frigate setup has been successfully initialized! 💡 Access it using the following URL: 🌐 http://192.168.1.9:5000`

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 11, 2025):

I just tried the default install in a 13th gen i5 Intel NUC with a Coral.ai. Proxmox 8.04. This machine was previously running frigate 0.15.

It completed but then goes in to a crashing loop. The web page comes up but I get a busy symbol. Sometimes with enough refreshes you see some content. I noticed that background task seem to be crashing and restarting. Eg. With a ps -aux I see these appear and disappear.

root 1503479 0.0 0.0 6940 2176 ? Ss 08:00 0:00 /bin/bash -c bash /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run 2> >(/usr/bin
root 1503480 0.0 0.0 6940 2304 ? S 08:00 0:00 bash /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run
root 1503481 9.0 0.0 13788 6912 ? S 08:00 0:00 /usr/bin/perl /usr/bin/ts %Y-%m-%d %H:%M:%.S
root 1503482 9.0 0.0 13752 6912 ? S 08:00 0:00 /usr/bin/perl /usr/bin/ts %Y-%m-%d %H:%M:%.S
root 1503487 140 0.0 197360 20992 ? Rl 08:00 0:00 python3 /usr/local/ffmpeg/get_ffmpeg_path.py

Tried again with an advanced install. Still crashing. Found this in the log file:

/opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run: line 20: /usr/lib/ffmpeg/7.0/bin/ffmpeg: No such file or directory

Within the lxc, I did this:

apt install -y ffmpeg
apt install -y vainfo intel-media-va-driver i965-va-driver
mkdir -p /usr/lib/ffmpeg/7.0/bin
ln -s /usr/bin/ffmpeg /usr/lib/ffmpeg/7.0/bin/ffmpeg
ln -s /usr/bin/ffprobe /usr/lib/ffmpeg/7.0/bin/ffprobe

Frigate is now running. Sample file still points to the wrong mp4 as I reported earlier.

Tried swapping out the openvino detector with:
coral:
type: edgetpu
device: pci

Back to crash cycling.

Part of log:

2025-09-11 08:38:12.985565 [2025-09-11 08:38:12] frigate.app INFO : go2rtc process pid: 3477
75
2025-09-11 08:38:12.998092 [2025-09-11 08:38:12] detector.coral INFO : Starting detection procc
ess: 37058
2025-09-11 08:38:12.999882 [2025-09-11 08:38:12] frigate.detectors.plugins.edgetpu_tfl INFO : Attempting to loo
ad TPU as pci
2025-09-11 08:38:13.001877 Process detector🪸
2025-09-11 08:38:13.001925 Traceback (most recent call last):
2025-09-11 08:38:13.001948 File "/usr/lib/python3.11/multiprocessing/process.py", line 314, in _bootstrap
2025-09-11 08:38:13.001965 self.run()
2025-09-11 08:38:13.001980 File "/opt/frigate/frigate/util/process.py", line 41, in run_wrapper
2025-09-11 08:38:13.001995 return run(*args, **kwargs)
2025-09-11 08:38:13.002011 ^^^^^^^^^^^^^^^^^^^^

The lxc conf file still has the extra slash on the export. Is this OK? I removed them and restarted the LXC

lxc.mount.entry: /dev/dri /dev/dri none bind,optional,create=dir
lxc.mount.entry: /dev/dri/renderD128 /dev/dri/renderD128 none bind,optional,create=file

I then did this (from earlier post):

echo "deb https://packages.cloud.google.com/apt coral-edgetpu-stable main"
| tee /etc/apt/sources.list.d/coral-edgetpu.list
curl https://packages.cloud.google.com/apt/doc/apt-key.gpg | apt-key add -

apt install -y libedgetpu1-std
apt install -y vainfo intel-media-va-driver i965-va-driver

But got this:

% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0Warning: apt-key is deprecated. Manage keyring files in trusted.gpg.d instead (see apt-key(8)).
100 1022 100 1022 0 0 5595 0 --:--:-- --:--:-- --:--:-- 5615
OK
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
E: Unable to locate package libedgetpu1-std
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
vainfo is already the newest version (2.12.0+ds1-1).
intel-media-va-driver is already the newest version (23.1.1+dfsg1-1).
i965-va-driver is already the newest version (2.4.1+dfsg1-1).
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.

I also attempted to install it on a N150 system with no extra decoders running Promox 9.0.x

Once again had to do:
apt install -y ffmpeg
apt install -y vainfo intel-media-va-driver i965-va-driver
mkdir -p /usr/lib/ffmpeg/7.0/bin
ln -s /usr/bin/ffmpeg /usr/lib/ffmpeg/7.0/bin/ffmpeg
ln -s /usr/bin/ffprobe /usr/lib/ffmpeg/7.0/bin/ffprobe

And correct the sample mp4 to person-bicycle-car-detection.mp4

It seems to be working. CPU usage is very high. I changed CPU to GPU in the config and got this:

2025-09-11 08:59:27.713072 2025-09-11 08:59:27.731972 2025-09-11 08:59:27.732051 2025-09-11 08:59:27.732076 2025-09-11 08:59:27.732095 2025-09-11 08:59:27.732114 2025-09-11 08:59:27.732134 2025-09-11 08:59:27.732153 2025-09-11 08:59:27.732173 2025-09-11 08:59:27.732193 2025-09-11 08:59:27.732212 2025-09-11 08:59:27.732232 2025-09-11 08:59:27.732252 2025-09-11 08:59:27.732273 2025-09-11 08:59:27.732293 2025-09-11 08:59:27.732314 2025-09-11 08:59:27.732335 2025-09-11 08:59:27.732356 2025-09-11 08:59:27.732377 2025-09-11 08:59:27.732397 2025-09-11 08:59:27.732416 2025-09-11 08:59:27.732435 2025-09-11 08:59:27.732454 in compile_model
2025-09-11 08:59:27.732473 2025-09-11 08:59:27.732492 2025-09-11 08:59:27.732511 2025-09-11 08:59:27.732531 2025-09-11 08:59:27.732550 2025-09-11 08:59:27.732570 2025-09-11 08:59:27.732605
2025-09-11 08:59:27.732630 [2025-09-11 08:59:27] frigate.app INFO : Camera processor started for test: 48277
Process detector:ov:
Traceback (most recent call last):
File "/usr/lib/python3.11/multiprocessing/process.py", line 314, in _bootstrap
self.run()
File "/opt/frigate/frigate/util/process.py", line 41, in run_wrapper
return run(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^
File "/usr/lib/python3.11/multiprocessing/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/opt/frigate/frigate/object_detection/base.py", line 112, in run_detector
object_detector = LocalObjectDetector(detector_config=detector_config)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/frigate/frigate/object_detection/base.py", line 57, in init
self.detect_api = create_detector(detector_config)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/frigate/frigate/detectors/init.py", line 18, in create_detector
return api(detector_config)
^^^^^^^^^^^^^^^^^^^^
File "/opt/frigate/frigate/detectors/plugins/openvino.py", line 50, in init
self.interpreter = self.ov_core.compile_model(
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/dist-packages/openvino/runtime/ie_api.py", line 543,
super().compile_model(model, device_name, {} if config is None else config),
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: Exception from src/inference/src/cpp/core.cpp:121:
Exception from src/inference/src/dev/plugin.cpp:58:
Check 'contexts.count(device_id)' failed at src/plugins/intel_gpu/src/plugin/plugin.cpp:232:
[GPU] Context was not initialized for 0 device

Again I edited the lxc conf file to remove the extra (?) slashes.

lxc.mount.entry: /dev/dri dev/dri none bind,optional,create=dir
lxc.cgroup2.devices.allow: c 0:0 rwm
lxc.mount.entry: /dev/dri/renderD128 dev/dri/renderD128 none bind,optional,create=file
lxc.cgroup2.devices.allow: c 0:0 rwm
lxc.mount.entry: /dev/dri/card0 dev/dri/card0 none bind,optional,create=file

Got this again: 2025-09-11 09:09:23.488006 [2025-09-11 09:09:23] frigate.app INFO : Capture process started
for test: 1095
2025-09-11 09:09:23.492069 Process detector:ov:
2025-09-11 09:09:23.492134 Traceback (most recent call last):
2025-09-11 09:09:23.492158 File "/usr/lib/python3.11/multiprocessing/process.py", line 314, in _bootstrap
2025-09-11 09:09:23.492177 self.run()
2025-09-11 09:09:23.492195 File "/opt/frigate/frigate/util/process.py", line 41, in run_wrapper
2025-09-11 09:09:23.492213 return run(*args, **kwargs)
2025-09-11 09:09:23.492231 ^^^^^^^^^^^^^^^^^^^^
2025-09-11 09:09:23.492249 File "/usr/lib/python3.11/multiprocessing/process.py", line 108, in run
2025-09-11 09:09:23.492267 self._target(*self._args, **self._kwargs)
2025-09-11 09:09:23.492285 File "/opt/frigate/frigate/object_detection/base.py", line 112, in run_detector
2025-09-11 09:09:23.492302 object_detector = LocalObjectDetector(detector_config=detector_config)
2025-09-11 09:09:23.492320 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2025-09-11 09:09:23.492337 File "/opt/frigate/frigate/object_detection/base.py", line 57, in init
2025-09-11 09:09:23.492355 self.detect_api = create_detector(detector_config)
2025-09-11 09:09:23.492372 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2025-09-11 09:09:23.492390 File "/opt/frigate/frigate/detectors/init.py", line 18, in create_detector
2025-09-11 09:09:23.492408 return api(detector_config)
2025-09-11 09:09:23.492426 ^^^^^^^^^^^^^^^^^^^^
2025-09-11 09:09:23.492444 File "/opt/frigate/frigate/detectors/plugins/openvino.py", line 50, in init
2025-09-11 09:09:23.492462 self.interpreter = self.ov_core.compile_model(
2025-09-11 09:09:23.492480 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
2025-09-11 09:09:23.492498 File "/usr/local/lib/python3.11/dist-packages/openvino/runtime/ie_api.py", line 543,
in compile_model
2025-09-11 09:09:23.492516 super().compile_model(model, device_name, {} if config is None else config),
2025-09-11 09:09:23.492534 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2025-09-11 09:09:23.492552 RuntimeError: Exception from src/inference/src/cpp/core.cpp:121:
2025-09-11 09:09:23.492570 Exception from src/inference/src/dev/plugin.cpp:58:
2025-09-11 09:09:23.492588 Check 'contexts.count(device_id)' failed at src/plugins/intel_gpu/src/plugin/plugin.cp
p:232:
2025-09-11 09:09:23.492625 [GPU] Context was not initialized for 0 device
2025-09-11 09:09:23.492646

Not sure if it was there before but earlier on this run I also got this in the log:

2025-09-11 09:09:23.396975 [2025-09-11 09:09:23] frigate.util.services WARNING : Did not detect hwaccel,
using a GPU for accelerated video decoding is highly recommended

I tried adding a line in the ffmpeg section of the frigate config:
hwaccel_args: preset-vaapi

Didn't help.

So I tried to run the intel_gpu_top command but first had to install it:
sudo apt install intel-gpu-tools

Got this upon running it:

intel-gpu-top: 8086:46d4 @ /dev/dri/card0 - 0/ 0 MHz; 100% RC6; 0 irqs/s

     ENGINES     BUSY                                                                         MI_SEMA MI_WAIT
   Render/3D    0.00% |                                                                     |      0%      0%
     Blitter    0.00% |                                                                     |      0%      0%
       Video    0.00% |                                                                     |      0%      0%
VideoEnhance    0.00% |                                                                     |      0%      0%

PID NAME Render/3D Blitter Video VideoEnhance

Really not sure what that means. Does it mean it's not seeing any hardware, or that it is not being used?

however:
/dev/shm/logs/frigate# intel_gpu_top -L
card0 8086:46d4 pci:vendor=8086,device=46D4,card=0
└─renderD128

So it seems like it may be there.

Put the hwaccel_args: preset-vaapi at the root level instead of in the camera and it seems to have stopped complaining about missing hwaccel

however, still failed:

2025-09-11 09:46:37.058525 [2025-09-11 09:46:37] frigate.app INFO : Capture process started
for test: 4093
2025-09-11 09:46:37.070149 Process detector:ov:
2025-09-11 09:46:37.070225 Traceback (most recent call last):
2025-09-11 09:46:37.070250 File "/usr/lib/python3.11/multiprocessing/process.py", line 314, in _bootstrap
2025-09-11 09:46:37.070269 self.run()
2025-09-11 09:46:37.070287 File "/opt/frigate/frigate/util/process.py", line 41, in run_wrapper
2025-09-11 09:46:37.070305 return run(*args, **kwargs)
2025-09-11 09:46:37.070323 ^^^^^^^^^^^^^^^^^^^^
---cut---

@atrozgrifo commented on GitHub (Sep 11, 2025): I just tried the default install in a 13th gen i5 Intel NUC with a Coral.ai. Proxmox 8.04. This machine was previously running frigate 0.15. It completed but then goes in to a crashing loop. The web page comes up but I get a busy symbol. Sometimes with enough refreshes you see some content. I noticed that background task seem to be crashing and restarting. Eg. With a ps -aux I see these appear and disappear. root 1503479 0.0 0.0 6940 2176 ? Ss 08:00 0:00 /bin/bash -c bash /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run 2> >(/usr/bin root 1503480 0.0 0.0 6940 2304 ? S 08:00 0:00 bash /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run root 1503481 9.0 0.0 13788 6912 ? S 08:00 0:00 /usr/bin/perl /usr/bin/ts %Y-%m-%d %H:%M:%.S root 1503482 9.0 0.0 13752 6912 ? S 08:00 0:00 /usr/bin/perl /usr/bin/ts %Y-%m-%d %H:%M:%.S root 1503487 140 0.0 197360 20992 ? Rl 08:00 0:00 python3 /usr/local/ffmpeg/get_ffmpeg_path.py Tried again with an advanced install. Still crashing. Found this in the log file: /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run: line 20: /usr/lib/ffmpeg/7.0/bin/ffmpeg: No such file or directory Within the lxc, I did this: apt install -y ffmpeg apt install -y vainfo intel-media-va-driver i965-va-driver mkdir -p /usr/lib/ffmpeg/7.0/bin ln -s /usr/bin/ffmpeg /usr/lib/ffmpeg/7.0/bin/ffmpeg ln -s /usr/bin/ffprobe /usr/lib/ffmpeg/7.0/bin/ffprobe Frigate is now running. Sample file still points to the wrong mp4 as I reported earlier. Tried swapping out the openvino detector with: coral: type: edgetpu device: pci Back to crash cycling. Part of log: 2025-09-11 08:38:12.985565 [2025-09-11 08:38:12] frigate.app INFO : go2rtc process pid: 3477 75 2025-09-11 08:38:12.998092 [2025-09-11 08:38:12] detector.coral INFO : Starting detection procc ess: 37058 2025-09-11 08:38:12.999882 [2025-09-11 08:38:12] frigate.detectors.plugins.edgetpu_tfl INFO : Attempting to loo ad TPU as pci 2025-09-11 08:38:13.001877 Process detector:coral: 2025-09-11 08:38:13.001925 Traceback (most recent call last): 2025-09-11 08:38:13.001948 File "/usr/lib/python3.11/multiprocessing/process.py", line 314, in _bootstrap 2025-09-11 08:38:13.001965 self.run() 2025-09-11 08:38:13.001980 File "/opt/frigate/frigate/util/process.py", line 41, in run_wrapper 2025-09-11 08:38:13.001995 return run(*args, **kwargs) 2025-09-11 08:38:13.002011 ^^^^^^^^^^^^^^^^^^^^ The lxc conf file still has the extra slash on the export. Is this OK? I removed them and restarted the LXC lxc.mount.entry: /dev/dri /dev/dri none bind,optional,create=dir lxc.mount.entry: /dev/dri/renderD128 /dev/dri/renderD128 none bind,optional,create=file I then did this (from earlier post): echo "deb https://packages.cloud.google.com/apt coral-edgetpu-stable main" \ | tee /etc/apt/sources.list.d/coral-edgetpu.list curl https://packages.cloud.google.com/apt/doc/apt-key.gpg | apt-key add - apt install -y libedgetpu1-std apt install -y vainfo intel-media-va-driver i965-va-driver But got this: % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0Warning: apt-key is deprecated. Manage keyring files in trusted.gpg.d instead (see apt-key(8)). 100 1022 100 1022 0 0 5595 0 --:--:-- --:--:-- --:--:-- 5615 OK Reading package lists... Done Building dependency tree... Done Reading state information... Done E: Unable to locate package libedgetpu1-std Reading package lists... Done Building dependency tree... Done Reading state information... Done vainfo is already the newest version (2.12.0+ds1-1). intel-media-va-driver is already the newest version (23.1.1+dfsg1-1). i965-va-driver is already the newest version (2.4.1+dfsg1-1). 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. I also attempted to install it on a N150 system with no extra decoders running Promox 9.0.x Once again had to do: apt install -y ffmpeg apt install -y vainfo intel-media-va-driver i965-va-driver mkdir -p /usr/lib/ffmpeg/7.0/bin ln -s /usr/bin/ffmpeg /usr/lib/ffmpeg/7.0/bin/ffmpeg ln -s /usr/bin/ffprobe /usr/lib/ffmpeg/7.0/bin/ffprobe And correct the sample mp4 to person-bicycle-car-detection.mp4 It seems to be working. CPU usage is very high. I changed CPU to GPU in the config and got this: 2025-09-11 08:59:27.713072 [2025-09-11 08:59:27] frigate.app INFO : Camera processor started for test: 48277 2025-09-11 08:59:27.731972 Process detector:ov: 2025-09-11 08:59:27.732051 Traceback (most recent call last): 2025-09-11 08:59:27.732076 File "/usr/lib/python3.11/multiprocessing/process.py", line 314, in _bootstrap 2025-09-11 08:59:27.732095 self.run() 2025-09-11 08:59:27.732114 File "/opt/frigate/frigate/util/process.py", line 41, in run_wrapper 2025-09-11 08:59:27.732134 return run(*args, **kwargs) 2025-09-11 08:59:27.732153 ^^^^^^^^^^^^^^^^^^^^ 2025-09-11 08:59:27.732173 File "/usr/lib/python3.11/multiprocessing/process.py", line 108, in run 2025-09-11 08:59:27.732193 self._target(*self._args, **self._kwargs) 2025-09-11 08:59:27.732212 File "/opt/frigate/frigate/object_detection/base.py", line 112, in run_detector 2025-09-11 08:59:27.732232 object_detector = LocalObjectDetector(detector_config=detector_config) 2025-09-11 08:59:27.732252 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 2025-09-11 08:59:27.732273 File "/opt/frigate/frigate/object_detection/base.py", line 57, in __init__ 2025-09-11 08:59:27.732293 self.detect_api = create_detector(detector_config) 2025-09-11 08:59:27.732314 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 2025-09-11 08:59:27.732335 File "/opt/frigate/frigate/detectors/__init__.py", line 18, in create_detector 2025-09-11 08:59:27.732356 return api(detector_config) 2025-09-11 08:59:27.732377 ^^^^^^^^^^^^^^^^^^^^ 2025-09-11 08:59:27.732397 File "/opt/frigate/frigate/detectors/plugins/openvino.py", line 50, in __init__ 2025-09-11 08:59:27.732416 self.interpreter = self.ov_core.compile_model( 2025-09-11 08:59:27.732435 ^^^^^^^^^^^^^^^^^^^^^^^^^^^ 2025-09-11 08:59:27.732454 File "/usr/local/lib/python3.11/dist-packages/openvino/runtime/ie_api.py", line 543, in compile_model 2025-09-11 08:59:27.732473 super().compile_model(model, device_name, {} if config is None else config), 2025-09-11 08:59:27.732492 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 2025-09-11 08:59:27.732511 RuntimeError: Exception from src/inference/src/cpp/core.cpp:121: 2025-09-11 08:59:27.732531 Exception from src/inference/src/dev/plugin.cpp:58: 2025-09-11 08:59:27.732550 Check 'contexts.count(device_id)' failed at src/plugins/intel_gpu/src/plugin/plugin.cpp:232: 2025-09-11 08:59:27.732570 [GPU] Context was not initialized for 0 device 2025-09-11 08:59:27.732605 2025-09-11 08:59:27.732630 Again I edited the lxc conf file to remove the extra (?) slashes. lxc.mount.entry: /dev/dri dev/dri none bind,optional,create=dir lxc.cgroup2.devices.allow: c 0:0 rwm lxc.mount.entry: /dev/dri/renderD128 dev/dri/renderD128 none bind,optional,create=file lxc.cgroup2.devices.allow: c 0:0 rwm lxc.mount.entry: /dev/dri/card0 dev/dri/card0 none bind,optional,create=file Got this again: 2025-09-11 09:09:23.488006 [2025-09-11 09:09:23] frigate.app INFO : Capture process started for test: 1095 2025-09-11 09:09:23.492069 Process detector:ov: 2025-09-11 09:09:23.492134 Traceback (most recent call last): 2025-09-11 09:09:23.492158 File "/usr/lib/python3.11/multiprocessing/process.py", line 314, in _bootstrap 2025-09-11 09:09:23.492177 self.run() 2025-09-11 09:09:23.492195 File "/opt/frigate/frigate/util/process.py", line 41, in run_wrapper 2025-09-11 09:09:23.492213 return run(*args, **kwargs) 2025-09-11 09:09:23.492231 ^^^^^^^^^^^^^^^^^^^^ 2025-09-11 09:09:23.492249 File "/usr/lib/python3.11/multiprocessing/process.py", line 108, in run 2025-09-11 09:09:23.492267 self._target(*self._args, **self._kwargs) 2025-09-11 09:09:23.492285 File "/opt/frigate/frigate/object_detection/base.py", line 112, in run_detector 2025-09-11 09:09:23.492302 object_detector = LocalObjectDetector(detector_config=detector_config) 2025-09-11 09:09:23.492320 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 2025-09-11 09:09:23.492337 File "/opt/frigate/frigate/object_detection/base.py", line 57, in __init__ 2025-09-11 09:09:23.492355 self.detect_api = create_detector(detector_config) 2025-09-11 09:09:23.492372 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 2025-09-11 09:09:23.492390 File "/opt/frigate/frigate/detectors/__init__.py", line 18, in create_detector 2025-09-11 09:09:23.492408 return api(detector_config) 2025-09-11 09:09:23.492426 ^^^^^^^^^^^^^^^^^^^^ 2025-09-11 09:09:23.492444 File "/opt/frigate/frigate/detectors/plugins/openvino.py", line 50, in __init__ 2025-09-11 09:09:23.492462 self.interpreter = self.ov_core.compile_model( 2025-09-11 09:09:23.492480 ^^^^^^^^^^^^^^^^^^^^^^^^^^^ 2025-09-11 09:09:23.492498 File "/usr/local/lib/python3.11/dist-packages/openvino/runtime/ie_api.py", line 543, in compile_model 2025-09-11 09:09:23.492516 super().compile_model(model, device_name, {} if config is None else config), 2025-09-11 09:09:23.492534 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 2025-09-11 09:09:23.492552 RuntimeError: Exception from src/inference/src/cpp/core.cpp:121: 2025-09-11 09:09:23.492570 Exception from src/inference/src/dev/plugin.cpp:58: 2025-09-11 09:09:23.492588 Check 'contexts.count(device_id)' failed at src/plugins/intel_gpu/src/plugin/plugin.cp p:232: 2025-09-11 09:09:23.492625 [GPU] Context was not initialized for 0 device 2025-09-11 09:09:23.492646 Not sure if it was there before but earlier on this run I also got this in the log: 2025-09-11 09:09:23.396975 [2025-09-11 09:09:23] frigate.util.services WARNING : Did not detect hwaccel, using a GPU for accelerated video decoding is highly recommended I tried adding a line in the ffmpeg section of the frigate config: hwaccel_args: preset-vaapi Didn't help. So I tried to run the intel_gpu_top command but first had to install it: sudo apt install intel-gpu-tools Got this upon running it: intel-gpu-top: 8086:46d4 @ /dev/dri/card0 - 0/ 0 MHz; 100% RC6; 0 irqs/s ENGINES BUSY MI_SEMA MI_WAIT Render/3D 0.00% | | 0% 0% Blitter 0.00% | | 0% 0% Video 0.00% | | 0% 0% VideoEnhance 0.00% | | 0% 0% PID NAME Render/3D Blitter Video VideoEnhance Really not sure what that means. Does it mean it's not seeing any hardware, or that it is not being used? however: /dev/shm/logs/frigate# intel_gpu_top -L card0 8086:46d4 pci:vendor=8086,device=46D4,card=0 └─renderD128 So it seems like it may be there. Put the hwaccel_args: preset-vaapi at the root level instead of in the camera and it seems to have stopped complaining about missing hwaccel however, still failed: 2025-09-11 09:46:37.058525 [2025-09-11 09:46:37] frigate.app INFO : Capture process started for test: 4093 2025-09-11 09:46:37.070149 Process detector:ov: 2025-09-11 09:46:37.070225 Traceback (most recent call last): 2025-09-11 09:46:37.070250 File "/usr/lib/python3.11/multiprocessing/process.py", line 314, in _bootstrap 2025-09-11 09:46:37.070269 self.run() 2025-09-11 09:46:37.070287 File "/opt/frigate/frigate/util/process.py", line 41, in run_wrapper 2025-09-11 09:46:37.070305 return run(*args, **kwargs) 2025-09-11 09:46:37.070323 ^^^^^^^^^^^^^^^^^^^^ ---cut---

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 11, 2025):

Sorry everyone, I was tired and pushed to the wrong branch. Please use this to test:
bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVED/raw/main/ct/frigate.sh)"

There is an issue with iGPU passthrough (which is hard for me to debug since I don't have one). For some reason, although the script mounts /dev/dri in the lxc config, the folder is not mounted inside the lxc. @MickLesk could you help investigate? If I build the LXC using my fork (bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz1337/ct/frigate16.sh)"), /dev/dri is present inside the LXC, so I'm thinking there's something different in the build.func of the community repo (my fork is a bit different and not up to date)

# USB passthrough
arch: amd64
cores: 4
features: nesting=1,fuse=1
hostname: frigate
memory: 4096
net0: name=eth0,bridge=vmbr0,hwaddr=BC:24:11:EB:76:02,ip=dhcp,type=veth
onboot: 1
ostype: debian
rootfs: local-lvm:vm-140-disk-0,size=20G
swap: 512
tags: community-script;nvr
lxc.cgroup2.devices.allow: a
lxc.cap.drop:
lxc.cgroup2.devices.allow: c 188:* rwm
lxc.cgroup2.devices.allow: c 189:* rwm
lxc.mount.entry: /dev/serial/by-id  dev/serial/by-id  none bind,optional,create=dir
lxc.mount.entry: /dev/ttyUSB0       dev/ttyUSB0       none bind,optional,create=file
lxc.mount.entry: /dev/ttyUSB1       dev/ttyUSB1       none bind,optional,create=file
lxc.mount.entry: /dev/ttyACM0       dev/ttyACM0       none bind,optional,create=file
lxc.mount.entry: /dev/ttyACM1       dev/ttyACM1       none bind,optional,create=file
lxc.mount.entry: /dev/dri /dev/dri none bind,optional,create=dir
lxc.cgroup2.devices.allow: c 226:128 rwm
lxc.mount.entry: /dev/dri/renderD128 /dev/dri/renderD128 none bind,optional,create=file
lxc.cgroup2.devices.allow: c 226:0 rwm
lxc.mount.entry: /dev/dri/card0 /dev/dri/card0 none bind,optional,create=file
lxc.cgroup2.devices.allow: c 10:200 rwm
lxc.mount.entry: /dev/net/tun dev/net/tun none bind,create=file

@remz1337 commented on GitHub (Sep 11, 2025): Sorry everyone, I was tired and pushed to the wrong branch. Please use this to test: `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVED/raw/main/ct/frigate.sh)"` There is an issue with iGPU passthrough (which is hard for me to debug since I don't have one). For some reason, although the script mounts `/dev/dri` in the lxc config, the folder is not mounted inside the lxc. @MickLesk could you help investigate? If I build the LXC using my fork (`bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz1337/ct/frigate16.sh)"`), `/dev/dri` is present inside the LXC, so I'm thinking there's something different in the `build.func` of the community repo (my fork is a bit different and not up to date) ``` # USB passthrough arch: amd64 cores: 4 features: nesting=1,fuse=1 hostname: frigate memory: 4096 net0: name=eth0,bridge=vmbr0,hwaddr=BC:24:11:EB:76:02,ip=dhcp,type=veth onboot: 1 ostype: debian rootfs: local-lvm:vm-140-disk-0,size=20G swap: 512 tags: community-script;nvr lxc.cgroup2.devices.allow: a lxc.cap.drop: lxc.cgroup2.devices.allow: c 188:* rwm lxc.cgroup2.devices.allow: c 189:* rwm lxc.mount.entry: /dev/serial/by-id dev/serial/by-id none bind,optional,create=dir lxc.mount.entry: /dev/ttyUSB0 dev/ttyUSB0 none bind,optional,create=file lxc.mount.entry: /dev/ttyUSB1 dev/ttyUSB1 none bind,optional,create=file lxc.mount.entry: /dev/ttyACM0 dev/ttyACM0 none bind,optional,create=file lxc.mount.entry: /dev/ttyACM1 dev/ttyACM1 none bind,optional,create=file lxc.mount.entry: /dev/dri /dev/dri none bind,optional,create=dir lxc.cgroup2.devices.allow: c 226:128 rwm lxc.mount.entry: /dev/dri/renderD128 /dev/dri/renderD128 none bind,optional,create=file lxc.cgroup2.devices.allow: c 226:0 rwm lxc.mount.entry: /dev/dri/card0 /dev/dri/card0 none bind,optional,create=file lxc.cgroup2.devices.allow: c 10:200 rwm lxc.mount.entry: /dev/net/tun dev/net/tun none bind,create=file ```

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 11, 2025):

Sorry everyone, I was tired and pushed to the wrong branch. Please use this to test: bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVED/raw/main/ct/frigate.sh)"

There is an issue with iGPU passthrough (which is hard for me to debug since I don't have one). For some reason, although the script mounts /dev/dri in the lxc config, the folder is not mounted inside the lxc

I tried on both machines (i5 13th gen with Coral.AI and an n150). Both gave me the iGPU problem I think

chgrp: cannot access '/dev/dri': No such file or directory

[ERROR] in line 28: exit code 0: while executing command chgrp video /dev/dri

[ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)"

/dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required

I keep asking about the extra slashes on the dri lines in the lxc file and whether that matters. Your file above has them again. The Frigate.video site does not.

I removed the extra slash on the 3 lines, rebooted the LXC and was able to execute "chgrp video /dev/dri". Looking with 'ls -l' in the LXC it is there and has the right permissions.

As the script failed there originally frigate is not running.

@atrozgrifo commented on GitHub (Sep 11, 2025): > Sorry everyone, I was tired and pushed to the wrong branch. Please use this to test: `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVED/raw/main/ct/frigate.sh)"` > > There is an issue with iGPU passthrough (which is hard for me to debug since I don't have one). For some reason, although the script mounts `/dev/dri` in the lxc config, the folder is not mounted inside the lxc I tried on both machines (i5 13th gen with Coral.AI and an n150). Both gave me the iGPU problem I think chgrp: cannot access '/dev/dri': No such file or directory [ERROR] in line 28: exit code 0: while executing command chgrp video /dev/dri [ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)" /dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required I keep asking about the extra slashes on the dri lines in the lxc file and whether that matters. Your file above has them again. The Frigate.video site does not. I removed the extra slash on the 3 lines, rebooted the LXC and was able to execute "chgrp video /dev/dri". Looking with 'ls -l' in the LXC it is there and has the right permissions. As the script failed there originally frigate is not running.

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 12, 2025):

@atrozgrifo regarding the /dev/dri extra slashes, do you mean that this:

lxc.mount.entry: /dev/dri /dev/dri none bind,optional,create=dir
lxc.mount.entry: /dev/dri/renderD128 /dev/dri/renderD128 none bind,optional,create=file
lxc.mount.entry: /dev/dri/card0 /dev/dri/card0 none bind,optional,create=file

should instead be this?

lxc.mount.entry: /dev/dri dev/dri none bind,optional,create=dir
lxc.mount.entry: /dev/dri/renderD128 dev/dri/renderD128 none bind,optional,create=file
lxc.mount.entry: /dev/dri/card0 dev/dri/card0 none bind,optional,create=file

That would be weird because the former has worked fine for me. Would you mind sharing a bit more info about your setup (Hardware, proxmox version, special configuration...)?

@remz1337 commented on GitHub (Sep 12, 2025): @atrozgrifo regarding the /dev/dri extra slashes, do you mean that this: ``` lxc.mount.entry: /dev/dri /dev/dri none bind,optional,create=dir lxc.mount.entry: /dev/dri/renderD128 /dev/dri/renderD128 none bind,optional,create=file lxc.mount.entry: /dev/dri/card0 /dev/dri/card0 none bind,optional,create=file ``` should instead be this? ``` lxc.mount.entry: /dev/dri dev/dri none bind,optional,create=dir lxc.mount.entry: /dev/dri/renderD128 dev/dri/renderD128 none bind,optional,create=file lxc.mount.entry: /dev/dri/card0 dev/dri/card0 none bind,optional,create=file ``` That would be weird because the former has worked fine for me. Would you mind sharing a bit more info about your setup (Hardware, proxmox version, special configuration...)?

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 12, 2025):

@atrozgrifo regarding the /dev/dri extra slashes, do you mean that this:

lxc.mount.entry: /dev/dri /dev/dri none bind,optional,create=dir
lxc.mount.entry: /dev/dri/renderD128 /dev/dri/renderD128 none bind,optional,create=file
lxc.mount.entry: /dev/dri/card0 /dev/dri/card0 none bind,optional,create=file

should instead be this?

Yes.

lxc.mount.entry: /dev/dri dev/dri none bind,optional,create=dir
lxc.mount.entry: /dev/dri/renderD128 dev/dri/renderD128 none bind,optional,create=file
lxc.mount.entry: /dev/dri/card0 dev/dri/card0 none bind,optional,create=file
That would be weird because the former has worked fine for me. Would you mind sharing a bit more info about your setup (Hardware, proxmox version, special configuration...)?

I've provided this in past responses. 2 systems. 1 an Intel NUC i5 13th gen with a CORAL.AI M.2 running proxmox 8.0.4. The other is a basic N150 running Proxmox 9.0.6. I've been running the advanced configuration, giving it a password, location for the files, etc. I've not gotten any sort of prompt asking about GPUs, Coral.AI, etc. I've been trying to get the N150 system running with Frigate so I can go back and upgrade my 8.0.4 proxmox system.

On both machines I have no /dev/dri initially. Stop the LXC, edit the lxc .conf file to remove the slashes and start the LXC. I now have /dev/dri.

See Proxmox section for Frigate documentation on this: https://docs.frigate.video/frigate/installation/

intel_gpu_top on the i5 which has a running Frigate from an old TTECK helper-script shows this:

intel-gpu-top: Intel Alderlake_p (Gen12) @ /dev/dri/card0 - 390/1449 MHz; 48% RC6; 0.24/12.08 W; 280 irqs/s

     ENGINES     BUSY                                                                                                                            MI_SEMA MI_WAIT
   Render/3D    0.00% |                                                                                                                        |      0%      0%
     Blitter    0.00% |                                                                                                                        |      0%      0%
       Video    0.00% |                                                                                                                        |      0%      0%
VideoEnhance    0.00% |                                                                                                                        |      0%      0%

PID NAME Render/3D Blitter Video VideoEnhance
2323932 ffmpeg | || ||▊ || |
2344531 ffmpeg | || ||▌ || |
2317214 ffmpeg | || ||▍ || |
2316544 ffmpeg |▋ || ||▏ || |
2343979 ffmpeg | || ||▍ || |
2341097 ffmpeg | || ||▍ || |
2343176 ffmpeg | || ||▍ || |
2317248 ffmpeg | || ||▎ || |
2317263 ffmpeg | || ||▏ || |
2317239 ffmpeg | || ||▏ || |

Same on my N150 shows what I posted above (basically nothing, I'm assuming because nothing is using the GPU).

@atrozgrifo commented on GitHub (Sep 12, 2025): > [@atrozgrifo](https://github.com/atrozgrifo) regarding the /dev/dri extra slashes, do you mean that this: > > ``` > lxc.mount.entry: /dev/dri /dev/dri none bind,optional,create=dir > lxc.mount.entry: /dev/dri/renderD128 /dev/dri/renderD128 none bind,optional,create=file > lxc.mount.entry: /dev/dri/card0 /dev/dri/card0 none bind,optional,create=file > ``` > > should instead be this? Yes. > > ``` > lxc.mount.entry: /dev/dri dev/dri none bind,optional,create=dir > lxc.mount.entry: /dev/dri/renderD128 dev/dri/renderD128 none bind,optional,create=file > lxc.mount.entry: /dev/dri/card0 dev/dri/card0 none bind,optional,create=file > ``` > > That would be weird because the former has worked fine for me. Would you mind sharing a bit more info about your setup (Hardware, proxmox version, special configuration...)? I've provided this in past responses. 2 systems. 1 an Intel NUC i5 13th gen with a CORAL.AI M.2 running proxmox 8.0.4. The other is a basic N150 running Proxmox 9.0.6. I've been running the advanced configuration, giving it a password, location for the files, etc. I've not gotten any sort of prompt asking about GPUs, Coral.AI, etc. I've been trying to get the N150 system running with Frigate so I can go back and upgrade my 8.0.4 proxmox system. On both machines I have no /dev/dri initially. Stop the LXC, edit the lxc .conf file to remove the slashes and start the LXC. I now have /dev/dri. See Proxmox section for Frigate documentation on this: https://docs.frigate.video/frigate/installation/ intel_gpu_top on the i5 which has a running Frigate from an old TTECK helper-script shows this: intel-gpu-top: Intel Alderlake_p (Gen12) @ /dev/dri/card0 - 390/1449 MHz; 48% RC6; 0.24/12.08 W; 280 irqs/s ENGINES BUSY MI_SEMA MI_WAIT Render/3D 0.00% | | 0% 0% Blitter 0.00% | | 0% 0% Video 0.00% | | 0% 0% VideoEnhance 0.00% | | 0% 0% PID NAME Render/3D Blitter Video VideoEnhance 2323932 ffmpeg | || ||▊ || | 2344531 ffmpeg | || ||▌ || | 2317214 ffmpeg | || ||▍ || | 2316544 ffmpeg |▋ || ||▏ || | 2343979 ffmpeg | || ||▍ || | 2341097 ffmpeg | || ||▍ || | 2343176 ffmpeg | || ||▍ || | 2317248 ffmpeg | || ||▎ || | 2317263 ffmpeg | || ||▏ || | 2317239 ffmpeg | || ||▏ || | Same on my N150 shows what I posted above (basically nothing, I'm assuming because nothing is using the GPU).

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 12, 2025):

If that's the case, then we would need to change the hw passthrough lines in build.func, which would impact other scripts using hw passthrough.

@remz1337 commented on GitHub (Sep 12, 2025): If that's the case, then we would need to change the hw passthrough lines in `build.func`, which would impact other scripts using hw passthrough.

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 12, 2025):

If that's the case, then we would need to change the hw passthrough lines in build.func, which would impact other scripts using hw passthrough.

I don't know if 'build.func' was used or not but I have some other containers that have USB passed through that also do NOT have the extra slash. I'm pretty sure one was done just with the proxmox GUI setting the hardware for the LXC. So I don't know if removing the slash for everything would break anything or not.

@atrozgrifo commented on GitHub (Sep 12, 2025): > If that's the case, then we would need to change the hw passthrough lines in `build.func`, which would impact other scripts using hw passthrough. I don't know if 'build.func' was used or not but I have some other containers that have USB passed through that also do NOT have the extra slash. I'm pretty sure one was done just with the proxmox GUI setting the hardware for the LXC. So I don't know if removing the slash for everything would break anything or not.

kerem commented

Author

Owner

@simplemice commented on GitHub (Sep 12, 2025):

Sorry everyone, I was tired and pushed to the wrong branch. Please use this to test: bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVED/raw/main/ct/frigate.sh)"
There is an issue with iGPU passthrough (which is hard for me to debug since I don't have one). For some reason, although the script mounts /dev/dri in the lxc config, the folder is not mounted inside the lxc

I tried on both machines (i5 13th gen with Coral.AI and an n150). Both gave me the iGPU problem I think

chgrp: cannot access '/dev/dri': No such file or directory

[ERROR] in line 28: exit code 0: while executing command chgrp video /dev/dri

[ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)"

/dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required

I keep asking about the extra slashes on the dri lines in the lxc file and whether that matters. Your file above has them again. The Frigate.video site does not.

I removed the extra slash on the 3 lines, rebooted the LXC and was able to execute "chgrp video /dev/dri". Looking with 'ls -l' in the LXC it is there and has the right permissions.

As the script failed there originally frigate is not running.

have same error

`
✔️ Installed Dependencies
⠧ Setting Up Hardware Accelerationchgrp: cannot access '/dev/dri': No such file or directory
⠴ Setting Up Hardware Acceleration
[ERROR] in line 25: exit code 0: while executing command chgrp video /dev/dri

[ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)"

/dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required
root@pve:~#
`

@simplemice commented on GitHub (Sep 12, 2025): > > Sorry everyone, I was tired and pushed to the wrong branch. Please use this to test: `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVED/raw/main/ct/frigate.sh)"` > > There is an issue with iGPU passthrough (which is hard for me to debug since I don't have one). For some reason, although the script mounts `/dev/dri` in the lxc config, the folder is not mounted inside the lxc > > I tried on both machines (i5 13th gen with Coral.AI and an n150). Both gave me the iGPU problem I think > > chgrp: cannot access '/dev/dri': No such file or directory > > [ERROR] in line 28: exit code 0: while executing command chgrp video /dev/dri > > [ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL [https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh](https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/$%7Bvar_install%7D.sh))" > > /dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL [https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh](https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/$%7Bvar_install%7D.sh))": numeric argument required > > I keep asking about the extra slashes on the dri lines in the lxc file and whether that matters. Your file above has them again. The Frigate.video site does not. > > I removed the extra slash on the 3 lines, rebooted the LXC and was able to execute "chgrp video /dev/dri". Looking with 'ls -l' in the LXC it is there and has the right permissions. > > As the script failed there originally frigate is not running. have same error ` ✔️ Installed Dependencies ⠧ Setting Up Hardware Accelerationchgrp: cannot access '/dev/dri': No such file or directory ⠴ Setting Up Hardware Acceleration [ERROR] in line 25: exit code 0: while executing command chgrp video /dev/dri [ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)" /dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required root@pve:~# `

kerem commented

Author

Owner

@iz8mbw commented on GitHub (Sep 12, 2025):

Sorry everyone, I was tired and pushed to the wrong branch. Please use this to test: bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVED/raw/main/ct/frigate.sh)"

There is an issue with iGPU passthrough (which is hard for me to debug since I don't have one). For some reason, although the script mounts /dev/dri in the lxc config, the folder is not mounted inside the lxc. @MickLesk could you help investigate? If I build the LXC using my fork (bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz1337/ct/frigate16.sh)"), /dev/dri is present inside the LXC, so I'm thinking there's something different in the build.func of the community repo (my fork is a bit different and not up to date)
# USB passthrough
arch: amd64
cores: 4
features: nesting=1,fuse=1
hostname: frigate
memory: 4096
net0: name=eth0,bridge=vmbr0,hwaddr=BC:24:11:EB:76:02,ip=dhcp,type=veth
onboot: 1
ostype: debian
rootfs: local-lvm:vm-140-disk-0,size=20G
swap: 512
tags: community-script;nvr
lxc.cgroup2.devices.allow: a
lxc.cap.drop:
lxc.cgroup2.devices.allow: c 188:* rwm
lxc.cgroup2.devices.allow: c 189:* rwm
lxc.mount.entry: /dev/serial/by-id  dev/serial/by-id  none bind,optional,create=dir
lxc.mount.entry: /dev/ttyUSB0       dev/ttyUSB0       none bind,optional,create=file
lxc.mount.entry: /dev/ttyUSB1       dev/ttyUSB1       none bind,optional,create=file
lxc.mount.entry: /dev/ttyACM0       dev/ttyACM0       none bind,optional,create=file
lxc.mount.entry: /dev/ttyACM1       dev/ttyACM1       none bind,optional,create=file
lxc.mount.entry: /dev/dri /dev/dri none bind,optional,create=dir
lxc.cgroup2.devices.allow: c 226:128 rwm
lxc.mount.entry: /dev/dri/renderD128 /dev/dri/renderD128 none bind,optional,create=file
lxc.cgroup2.devices.allow: c 226:0 rwm
lxc.mount.entry: /dev/dri/card0 /dev/dri/card0 none bind,optional,create=file
lxc.cgroup2.devices.allow: c 10:200 rwm
lxc.mount.entry: /dev/net/tun dev/net/tun none bind,create=file

Just tested and script stop with this error:

chgrp: cannot access '/dev/dri': No such file or directory

[ERROR] in line 25: exit code 0: while executing command chgrp video /dev/dri

[ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)"

/dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required

@iz8mbw commented on GitHub (Sep 12, 2025): > Sorry everyone, I was tired and pushed to the wrong branch. Please use this to test: `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVED/raw/main/ct/frigate.sh)"` > > There is an issue with iGPU passthrough (which is hard for me to debug since I don't have one). For some reason, although the script mounts `/dev/dri` in the lxc config, the folder is not mounted inside the lxc. [@MickLesk](https://github.com/MickLesk) could you help investigate? If I build the LXC using my fork (`bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz1337/ct/frigate16.sh)"`), `/dev/dri` is present inside the LXC, so I'm thinking there's something different in the `build.func` of the community repo (my fork is a bit different and not up to date) > > ``` > # USB passthrough > arch: amd64 > cores: 4 > features: nesting=1,fuse=1 > hostname: frigate > memory: 4096 > net0: name=eth0,bridge=vmbr0,hwaddr=BC:24:11:EB:76:02,ip=dhcp,type=veth > onboot: 1 > ostype: debian > rootfs: local-lvm:vm-140-disk-0,size=20G > swap: 512 > tags: community-script;nvr > lxc.cgroup2.devices.allow: a > lxc.cap.drop: > lxc.cgroup2.devices.allow: c 188:* rwm > lxc.cgroup2.devices.allow: c 189:* rwm > lxc.mount.entry: /dev/serial/by-id dev/serial/by-id none bind,optional,create=dir > lxc.mount.entry: /dev/ttyUSB0 dev/ttyUSB0 none bind,optional,create=file > lxc.mount.entry: /dev/ttyUSB1 dev/ttyUSB1 none bind,optional,create=file > lxc.mount.entry: /dev/ttyACM0 dev/ttyACM0 none bind,optional,create=file > lxc.mount.entry: /dev/ttyACM1 dev/ttyACM1 none bind,optional,create=file > lxc.mount.entry: /dev/dri /dev/dri none bind,optional,create=dir > lxc.cgroup2.devices.allow: c 226:128 rwm > lxc.mount.entry: /dev/dri/renderD128 /dev/dri/renderD128 none bind,optional,create=file > lxc.cgroup2.devices.allow: c 226:0 rwm > lxc.mount.entry: /dev/dri/card0 /dev/dri/card0 none bind,optional,create=file > lxc.cgroup2.devices.allow: c 10:200 rwm > lxc.mount.entry: /dev/net/tun dev/net/tun none bind,create=file > ``` Just tested and script stop with this error: ``` chgrp: cannot access '/dev/dri': No such file or directory [ERROR] in line 25: exit code 0: while executing command chgrp video /dev/dri [ERROR] in line 1446: exit code 1: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)" /dev/fd/63: line 49: exit: lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/main/install/${var_install}.sh)": numeric argument required ```

kerem commented

Author

Owner

@iz8mbw commented on GitHub (Sep 12, 2025):

Don't know if this can help, this is the config file of a working (previous Frigate version) of the Frigate LXC created in the past with helper script:


# cat /etc/pve/nodes/pve01/lxc/103.conf 


arch: amd64
cores: 4
features: nesting=1
hostname: frigate01
memory: 5120
net0: name=eth0,bridge=vmbr0,hwaddr=BC:24:11:18:11:A3,ip=dhcp,type=veth
onboot: 1
ostype: debian
rootfs: nvme:vm-103-disk-0,size=110G
startup: up=75
swap: 512
tags: proxmox-helper-scripts
lxc.cgroup2.devices.allow: a
lxc.cap.drop: 
lxc.cgroup2.devices.allow: c 188:* rwm
lxc.cgroup2.devices.allow: c 189:* rwm
lxc.mount.entry: /dev/serial/by-id  dev/serial/by-id  none bind,optional,create=dir
lxc.mount.entry: /dev/ttyUSB0       dev/ttyUSB0       none bind,optional,create=file
lxc.mount.entry: /dev/ttyUSB1       dev/ttyUSB1       none bind,optional,create=file
lxc.mount.entry: /dev/ttyACM0       dev/ttyACM0       none bind,optional,create=file
lxc.mount.entry: /dev/ttyACM1       dev/ttyACM1       none bind,optional,create=file
lxc.cgroup2.devices.allow: c 226:0 rwm
lxc.cgroup2.devices.allow: c 226:128 rwm
lxc.cgroup2.devices.allow: c 29:0 rwm
lxc.mount.entry: /dev/fb0 dev/fb0 none bind,optional,create=file
lxc.mount.entry: /dev/dri dev/dri none bind,optional,create=dir
lxc.mount.entry: /dev/dri/renderD128 dev/dri/renderD128 none bind,optional,create=file

@iz8mbw commented on GitHub (Sep 12, 2025): Don't know if this can help, this is the config file of a working (previous Frigate version) of the Frigate LXC created in the past with helper script: ``` # cat /etc/pve/nodes/pve01/lxc/103.conf arch: amd64 cores: 4 features: nesting=1 hostname: frigate01 memory: 5120 net0: name=eth0,bridge=vmbr0,hwaddr=BC:24:11:18:11:A3,ip=dhcp,type=veth onboot: 1 ostype: debian rootfs: nvme:vm-103-disk-0,size=110G startup: up=75 swap: 512 tags: proxmox-helper-scripts lxc.cgroup2.devices.allow: a lxc.cap.drop: lxc.cgroup2.devices.allow: c 188:* rwm lxc.cgroup2.devices.allow: c 189:* rwm lxc.mount.entry: /dev/serial/by-id dev/serial/by-id none bind,optional,create=dir lxc.mount.entry: /dev/ttyUSB0 dev/ttyUSB0 none bind,optional,create=file lxc.mount.entry: /dev/ttyUSB1 dev/ttyUSB1 none bind,optional,create=file lxc.mount.entry: /dev/ttyACM0 dev/ttyACM0 none bind,optional,create=file lxc.mount.entry: /dev/ttyACM1 dev/ttyACM1 none bind,optional,create=file lxc.cgroup2.devices.allow: c 226:0 rwm lxc.cgroup2.devices.allow: c 226:128 rwm lxc.cgroup2.devices.allow: c 29:0 rwm lxc.mount.entry: /dev/fb0 dev/fb0 none bind,optional,create=file lxc.mount.entry: /dev/dri dev/dri none bind,optional,create=dir lxc.mount.entry: /dev/dri/renderD128 dev/dri/renderD128 none bind,optional,create=file ```

kerem commented

Author

Owner

@MickLesk commented on GitHub (Sep 12, 2025):

And the new one as compare?

@MickLesk commented on GitHub (Sep 12, 2025): And the new one as compare?

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 12, 2025):

@remz1337 look at the jellyfin script. It installs & runs just fine with an iGPU pve host.
And indeed, the mounts are without the slash, as @atrozgrifo mentioned a couple of times :

lxc.cgroup2.devices.allow: a
lxc.cap.drop: 
lxc.cgroup2.devices.allow: c 188:* rwm
lxc.cgroup2.devices.allow: c 189:* rwm
lxc.mount.entry: /dev/serial/by-id  dev/serial/by-id  none bind,optional,create=dir
lxc.mount.entry: /dev/ttyUSB0       dev/ttyUSB0       none bind,optional,create=file
lxc.mount.entry: /dev/ttyUSB1       dev/ttyUSB1       none bind,optional,create=file
lxc.mount.entry: /dev/ttyACM0       dev/ttyACM0       none bind,optional,create=file
lxc.mount.entry: /dev/ttyACM1       dev/ttyACM1       none bind,optional,create=file
lxc.cgroup2.devices.allow: c 226:0 rwm
lxc.cgroup2.devices.allow: c 226:128 rwm
lxc.cgroup2.devices.allow: c 29:0 rwm
lxc.mount.entry: /dev/fb0 dev/fb0 none bind,optional,create=file
lxc.mount.entry: /dev/dri dev/dri none bind,optional,create=dir
lxc.mount.entry: /dev/dri/renderD128 dev/dri/renderD128 none bind,optional,create=file

Official ProxmoxVE installs hw-acceleration without the leading slashes, you seem to be using https://github.com/remz1337/ProxmoxVED/blob/main/misc/build.func#L1259 , which is adding slashes :
echo "lxc.mount.entry: /dev/dri /dev/dri none bind,optional,create=dir" >>"$LXC_CONFIG"

@Zixim commented on GitHub (Sep 12, 2025): @remz1337 look at the jellyfin script. It installs & runs just fine with an iGPU pve host. And indeed, the mounts are without the slash, as @atrozgrifo mentioned a couple of times : ``` lxc.cgroup2.devices.allow: a lxc.cap.drop: lxc.cgroup2.devices.allow: c 188:* rwm lxc.cgroup2.devices.allow: c 189:* rwm lxc.mount.entry: /dev/serial/by-id dev/serial/by-id none bind,optional,create=dir lxc.mount.entry: /dev/ttyUSB0 dev/ttyUSB0 none bind,optional,create=file lxc.mount.entry: /dev/ttyUSB1 dev/ttyUSB1 none bind,optional,create=file lxc.mount.entry: /dev/ttyACM0 dev/ttyACM0 none bind,optional,create=file lxc.mount.entry: /dev/ttyACM1 dev/ttyACM1 none bind,optional,create=file lxc.cgroup2.devices.allow: c 226:0 rwm lxc.cgroup2.devices.allow: c 226:128 rwm lxc.cgroup2.devices.allow: c 29:0 rwm lxc.mount.entry: /dev/fb0 dev/fb0 none bind,optional,create=file lxc.mount.entry: /dev/dri dev/dri none bind,optional,create=dir lxc.mount.entry: /dev/dri/renderD128 dev/dri/renderD128 none bind,optional,create=file ``` Official ProxmoxVE installs hw-acceleration without the leading slashes, you seem to be using https://github.com/remz1337/ProxmoxVED/blob/main/misc/build.func#L1259 , which is adding slashes : `echo "lxc.mount.entry: /dev/dri /dev/dri none bind,optional,create=dir" >>"$LXC_CONFIG"`

kerem commented

Author

Owner

@iz8mbw commented on GitHub (Sep 12, 2025):

And the new one as compare?

root@pve01:~# cat /etc/pve/nodes/pve01/lxc/108.conf 
arch: amd64
cores: 4
features: nesting=1
hostname: fg16
memory: 4096
net0: name=eth0,bridge=vmbr0,hwaddr=BC:24:11:E6:51:9D,ip=dhcp,type=veth
onboot: 1
ostype: debian
rootfs: local-lvm:vm-108-disk-0,size=40G
swap: 512
tags: community-script;nvr
# USB passthrough
lxc.cgroup2.devices.allow: a
lxc.cap.drop:
lxc.cgroup2.devices.allow: c 188:* rwm
lxc.cgroup2.devices.allow: c 189:* rwm
lxc.mount.entry: /dev/serial/by-id  dev/serial/by-id  none bind,optional,create=dir
lxc.mount.entry: /dev/ttyUSB0       dev/ttyUSB0       none bind,optional,create=file
lxc.mount.entry: /dev/ttyUSB1       dev/ttyUSB1       none bind,optional,create=file
lxc.mount.entry: /dev/ttyACM0       dev/ttyACM0       none bind,optional,create=file
lxc.mount.entry: /dev/ttyACM1       dev/ttyACM1       none bind,optional,create=file
lxc.mount.entry: /dev/dri /dev/dri none bind,optional,create=dir
lxc.cgroup2.devices.allow: c 0:0 rwm
lxc.mount.entry: /dev/dri/renderD128 /dev/dri/renderD128 none bind,optional,create=file
lxc.cgroup2.devices.allow: c 0:0 rwm
lxc.mount.entry: /dev/dri/card0 /dev/dri/card0 none bind,optional,create=file

@iz8mbw commented on GitHub (Sep 12, 2025): > And the new one as compare? ``` root@pve01:~# cat /etc/pve/nodes/pve01/lxc/108.conf arch: amd64 cores: 4 features: nesting=1 hostname: fg16 memory: 4096 net0: name=eth0,bridge=vmbr0,hwaddr=BC:24:11:E6:51:9D,ip=dhcp,type=veth onboot: 1 ostype: debian rootfs: local-lvm:vm-108-disk-0,size=40G swap: 512 tags: community-script;nvr # USB passthrough lxc.cgroup2.devices.allow: a lxc.cap.drop: lxc.cgroup2.devices.allow: c 188:* rwm lxc.cgroup2.devices.allow: c 189:* rwm lxc.mount.entry: /dev/serial/by-id dev/serial/by-id none bind,optional,create=dir lxc.mount.entry: /dev/ttyUSB0 dev/ttyUSB0 none bind,optional,create=file lxc.mount.entry: /dev/ttyUSB1 dev/ttyUSB1 none bind,optional,create=file lxc.mount.entry: /dev/ttyACM0 dev/ttyACM0 none bind,optional,create=file lxc.mount.entry: /dev/ttyACM1 dev/ttyACM1 none bind,optional,create=file lxc.mount.entry: /dev/dri /dev/dri none bind,optional,create=dir lxc.cgroup2.devices.allow: c 0:0 rwm lxc.mount.entry: /dev/dri/renderD128 /dev/dri/renderD128 none bind,optional,create=file lxc.cgroup2.devices.allow: c 0:0 rwm lxc.mount.entry: /dev/dri/card0 /dev/dri/card0 none bind,optional,create=file ```

kerem commented

VE:
github.com/community-scripts/ProxmoxVE@aeb15c7ea1/misc/build.func (L1218)

Author

Owner

@remz1337 commented on GitHub (Sep 12, 2025):

Thanks for pointing it out. That's one of the reason why I hate the whole VE/VED split. Testing is a nightmare.

@MickLesk build.func in VED needs to be fixed to mount /dev/dri correctly. Can you help with that? In VE, it is mounted correctly, see comparison below:

VED:

echo "lxc.mount.entry: /dev/dri /dev/dri none bind,optional,create=dir" >>"$LXC_CONFIG"
github.com/community-scripts/ProxmoxVED@c20d542698/misc/build.func (L1784)

@remz1337 commented on GitHub (Sep 12, 2025): Thanks for pointing it out. That's one of the reason why I hate the whole VE/VED split. Testing is a nightmare. @MickLesk `build.func` in VED needs to be fixed to mount `/dev/dri` correctly. Can you help with that? In VE, it is mounted correctly, see comparison below: VE: https://github.com/community-scripts/ProxmoxVE/blob/aeb15c7ea1ffa719fd3e05af9028a9a28db2d0f6/misc/build.func#L1218 VED: ` echo "lxc.mount.entry: /dev/dri /dev/dri none bind,optional,create=dir" >>"$LXC_CONFIG"` https://github.com/community-scripts/ProxmoxVED/blob/c20d54269882079c5ed0a2a703f49f1d694ced9b/misc/build.func#L1784

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 12, 2025):

I have changed the source of the build.func to be the one on the community-repo. Might be worth to give it another go see if it works now

@remz1337 commented on GitHub (Sep 12, 2025): I have changed the source of the build.func to be the one on the community-repo. Might be worth to give it another go see if it works now

kerem commented

Author

Owner

@MickLesk commented on GitHub (Sep 12, 2025):

Thats an Feature Request by another, over ID instead of byName That doesnt Work correctly but i dont have get any Feedback anymore for this topic.

@MickLesk commented on GitHub (Sep 12, 2025): Thats an Feature Request by another, over ID instead of byName That doesnt Work correctly but i dont have get any Feedback anymore for this topic.

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 12, 2025):

@remz1337 Yes it installs w/o errors now.

However the frigate service is in a restart loop (which causes the infinite loading loop on the webpage) :

root@frigate:~# systemctl restart frigate ; journalctl -fu frigate
Sep 12 15:40:48 frigate systemd[1]: frigate.service: Deactivated successfully.
Sep 12 15:40:49 frigate systemd[1]: frigate.service: Scheduled restart job, restart counter is at 1.
Sep 12 15:40:49 frigate systemd[1]: Stopped frigate.service - Frigate service.
Sep 12 15:40:49 frigate systemd[1]: Starting frigate.service - Frigate service...
Sep 12 15:40:49 frigate systemd[1]: Started frigate.service - Frigate service.
Sep 12 15:40:50 frigate systemd[1]: frigate.service: Deactivated successfully.
Sep 12 15:40:51 frigate systemd[1]: frigate.service: Scheduled restart job, restart counter is at 2.
Sep 12 15:40:51 frigate systemd[1]: Stopped frigate.service - Frigate service.
Sep 12 15:40:51 frigate systemd[1]: Starting frigate.service - Frigate service...
Sep 12 15:40:51 frigate systemd[1]: Started frigate.service - Frigate service.
Sep 12 15:40:51 frigate systemd[1]: frigate.service: Deactivated successfully.
Sep 12 15:40:52 frigate systemd[1]: frigate.service: Scheduled restart job, restart counter is at 3.
Sep 12 15:40:52 frigate systemd[1]: Stopped frigate.service - Frigate service.
Sep 12 15:40:52 frigate systemd[1]: Starting frigate.service - Frigate service...
Sep 12 15:40:52 frigate systemd[1]: Started frigate.service - Frigate service.
Sep 12 15:40:53 frigate systemd[1]: frigate.service: Deactivated successfully.
...

This might be caused by the missing packages intel-media-driver , libva2 and vainfo as others have stated earlier in this thread.

update :
indeed, as @rommess wrote, these are required to get frigate up :

apt install -y ffmpeg
apt install -y vainfo intel-media-va-driver i965-va-driver
mkdir -p /usr/lib/ffmpeg/7.0/bin
ln -s /usr/bin/ffmpeg /usr/lib/ffmpeg/7.0/bin/ffmpeg
ln -s /usr/bin/ffprobe /usr/lib/ffmpeg/7.0/bin/ffprobe

More testing is needed but at least now we have a working installation path.

@MickLesk do you think apt install ffmpeg is right ?
Asking because this installs ffmpeg version 5.1.7-0+deb12u1, but Jellyfin with iGPU has a totally different version ffmpeg version 7.1.1-Jellyfin.

The Jellyfin-ffmpeg has extra flags set for HW acceleration, such as --enable-vaapi --enable-amf --enable-libvpl --enable-ffnvcodec --enable-cuda --enable-cuda-llvm --enable-cuvid --enable-nvdec --enable-nvenc.

Would it even make any kind of difference for Frigate ?

@Zixim commented on GitHub (Sep 12, 2025): @remz1337 Yes it installs w/o errors now. However the frigate service is in a restart loop (which causes the infinite loading loop on the webpage) : ``` root@frigate:~# systemctl restart frigate ; journalctl -fu frigate Sep 12 15:40:48 frigate systemd[1]: frigate.service: Deactivated successfully. Sep 12 15:40:49 frigate systemd[1]: frigate.service: Scheduled restart job, restart counter is at 1. Sep 12 15:40:49 frigate systemd[1]: Stopped frigate.service - Frigate service. Sep 12 15:40:49 frigate systemd[1]: Starting frigate.service - Frigate service... Sep 12 15:40:49 frigate systemd[1]: Started frigate.service - Frigate service. Sep 12 15:40:50 frigate systemd[1]: frigate.service: Deactivated successfully. Sep 12 15:40:51 frigate systemd[1]: frigate.service: Scheduled restart job, restart counter is at 2. Sep 12 15:40:51 frigate systemd[1]: Stopped frigate.service - Frigate service. Sep 12 15:40:51 frigate systemd[1]: Starting frigate.service - Frigate service... Sep 12 15:40:51 frigate systemd[1]: Started frigate.service - Frigate service. Sep 12 15:40:51 frigate systemd[1]: frigate.service: Deactivated successfully. Sep 12 15:40:52 frigate systemd[1]: frigate.service: Scheduled restart job, restart counter is at 3. Sep 12 15:40:52 frigate systemd[1]: Stopped frigate.service - Frigate service. Sep 12 15:40:52 frigate systemd[1]: Starting frigate.service - Frigate service... Sep 12 15:40:52 frigate systemd[1]: Started frigate.service - Frigate service. Sep 12 15:40:53 frigate systemd[1]: frigate.service: Deactivated successfully. ... ``` This might be caused by the missing packages `intel-media-driver , libva2 and vainfo` as others have stated earlier in this thread. update : indeed, as @rommess wrote, these are required to get frigate up : ``` apt install -y ffmpeg apt install -y vainfo intel-media-va-driver i965-va-driver mkdir -p /usr/lib/ffmpeg/7.0/bin ln -s /usr/bin/ffmpeg /usr/lib/ffmpeg/7.0/bin/ffmpeg ln -s /usr/bin/ffprobe /usr/lib/ffmpeg/7.0/bin/ffprobe ``` More testing is needed but at least now we have a working installation path. @MickLesk do you think `apt install ffmpeg` is right ? Asking because this installs `ffmpeg version 5.1.7-0+deb12u1`, but Jellyfin with iGPU has a totally different version `ffmpeg version 7.1.1-Jellyfin`. The Jellyfin-ffmpeg has extra flags set for HW acceleration, such as `--enable-vaapi --enable-amf --enable-libvpl --enable-ffnvcodec --enable-cuda --enable-cuda-llvm --enable-cuvid --enable-nvdec --enable-nvenc`. Would it even make any kind of difference for Frigate ?

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 12, 2025):

I have changed the source of the build.func to be the one on the community-repo. Might be worth to give it another go see if it works now

Tried on both systems, both still don't have ffmpeg in the right location.

2025-09-12 09:42:18.755397 /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run: line 20: /usr/lib/ffmpeg/7.0/bin/ffmpeg: No such file or directory

2025-09-12 09:43:57.871343 /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run: line 20: /usr/lib/ffmpeg/7.0/bin/ffmpeg: No such file or directory

Doing the following seems to correct this:

apt update
apt install -y ffmpeg

mkdir -p /usr/lib/ffmpeg/7.0/bin
ln -s /usr/bin/ffmpeg /usr/lib/ffmpeg/7.0/bin/ffmpeg
ln -s /usr/bin/ffprobe /usr/lib/ffmpeg/7.0/bin/ffprobe

Path to the sample mp4 still needs to be fixed.
path: /media/frigate/person-bicycle-car-detection.mp4

Now runs on both systems with CPU. Simply changing to GPU breaks it again.

2025-09-12 10:03:36.174960 [2025-09-12 10:03:36] frigate.app INFO : Camera processor started for test: 48785
2025-09-12 10:03:36.175003 Process detector:ov:
2025-09-12 10:03:36.175011 Traceback (most recent call last):
2025-09-12 10:03:36.175018 File "/usr/lib/python3.11/multiprocessing/process.py", line 314, in _bootstrap
2025-09-12 10:03:36.175025 self.run()
2025-09-12 10:03:36.175032 File "/opt/frigate/frigate/util/process.py", line 41, in run_wrapper
2025-09-12 10:03:36.175038 return run(*args, **kwargs)
2025-09-12 10:03:36.175045 ^^^^^^^^^^^^^^^^^^^^
2025-09-12 10:03:36.175052 File "/usr/lib/python3.11/multiprocessing/process.py", line 108, in run
2025-09-12 10:03:36.175059 self._target(*self._args, **self._kwargs)
2025-09-12 10:03:36.175065 File "/opt/frigate/frigate/object_detection/base.py", line 112, in run_detector
2025-09-12 10:03:36.175072 object_detector = LocalObjectDetector(detector_config=detector_config)
2025-09-12 10:03:36.175079 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2025-09-12 10:03:36.175086 File "/opt/frigate/frigate/object_detection/base.py", line 57, in init
2025-09-12 10:03:36.175092 self.detect_api = create_detector(detector_config)
2025-09-12 10:03:36.175099 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2025-09-12 10:03:36.175106 File "/opt/frigate/frigate/detectors/init.py", line 18, in create_detector
2025-09-12 10:03:36.175113 return api(detector_config)
2025-09-12 10:03:36.175120 ^^^^^^^^^^^^^^^^^^^^
2025-09-12 10:03:36.175126 File "/opt/frigate/frigate/detectors/plugins/openvino.py", line 50, in init
2025-09-12 10:03:36.175133 self.interpreter = self.ov_core.compile_model(
2025-09-12 10:03:36.175140 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
2025-09-12 10:03:36.175147 File "/usr/local/lib/python3.11/dist-packages/openvino/runtime/ie_api.py", line 543, in compile_model
2025-09-12 10:03:36.175154 super().compile_model(model, device_name, {} if config is None else config),
2025-09-12 10:03:36.175160 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2025-09-12 10:03:36.175167 RuntimeError: Exception from src/inference/src/cpp/core.cpp:121:
2025-09-12 10:03:36.175174 Exception from src/inference/src/dev/plugin.cpp:58:
2025-09-12 10:03:36.175180 Check 'contexts.count(device_id)' failed at src/plugins/intel_gpu/src/plugin/plugin.cpp:232:
2025-09-12 10:03:36.175187 [GPU] Context was not initialized for 0 device
2025-09-12 10:03:36.175198
2025-09-12 10:03:36.175205
2025-09-12 10:03:36.175211

Added to frigate config.yml in the camera section (didn't like it in the root).

ffmpeg:
hwaccel_args: preset-vaapi

Now gives a bunch of errors, I assume because ffmpeg fails because it can't get the hardware acceleration.

Neither system have a /dev/dri although I didn't see the chown error this time. Both lxc .conf files still have the extra /. Did I run the wrong command? I did get different results than the last time I ran this,.

One machine was:
bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/pr-frigate/ct/frigate.sh)"
The other was:
bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVED/raw/main/ct/frigate.sh)"

I cut and paste from above and I guess I grabbed different lines. Same result though.

@atrozgrifo commented on GitHub (Sep 12, 2025): > I have changed the source of the build.func to be the one on the community-repo. Might be worth to give it another go see if it works now Tried on both systems, both still don't have ffmpeg in the right location. 2025-09-12 09:42:18.755397 /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run: line 20: /usr/lib/ffmpeg/7.0/bin/ffmpeg: No such file or directory 2025-09-12 09:43:57.871343 /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run: line 20: /usr/lib/ffmpeg/7.0/bin/ffmpeg: No such file or directory Doing the following seems to correct this: apt update apt install -y ffmpeg mkdir -p /usr/lib/ffmpeg/7.0/bin ln -s /usr/bin/ffmpeg /usr/lib/ffmpeg/7.0/bin/ffmpeg ln -s /usr/bin/ffprobe /usr/lib/ffmpeg/7.0/bin/ffprobe Path to the sample mp4 still needs to be fixed. path: /media/frigate/person-bicycle-car-detection.mp4 Now runs on both systems with CPU. Simply changing to GPU breaks it again. 2025-09-12 10:03:36.174960 [2025-09-12 10:03:36] frigate.app INFO : Camera processor started for test: 48785 2025-09-12 10:03:36.175003 Process detector:ov: 2025-09-12 10:03:36.175011 Traceback (most recent call last): 2025-09-12 10:03:36.175018 File "/usr/lib/python3.11/multiprocessing/process.py", line 314, in _bootstrap 2025-09-12 10:03:36.175025 self.run() 2025-09-12 10:03:36.175032 File "/opt/frigate/frigate/util/process.py", line 41, in run_wrapper 2025-09-12 10:03:36.175038 return run(*args, **kwargs) 2025-09-12 10:03:36.175045 ^^^^^^^^^^^^^^^^^^^^ 2025-09-12 10:03:36.175052 File "/usr/lib/python3.11/multiprocessing/process.py", line 108, in run 2025-09-12 10:03:36.175059 self._target(*self._args, **self._kwargs) 2025-09-12 10:03:36.175065 File "/opt/frigate/frigate/object_detection/base.py", line 112, in run_detector 2025-09-12 10:03:36.175072 object_detector = LocalObjectDetector(detector_config=detector_config) 2025-09-12 10:03:36.175079 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 2025-09-12 10:03:36.175086 File "/opt/frigate/frigate/object_detection/base.py", line 57, in __init__ 2025-09-12 10:03:36.175092 self.detect_api = create_detector(detector_config) 2025-09-12 10:03:36.175099 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 2025-09-12 10:03:36.175106 File "/opt/frigate/frigate/detectors/__init__.py", line 18, in create_detector 2025-09-12 10:03:36.175113 return api(detector_config) 2025-09-12 10:03:36.175120 ^^^^^^^^^^^^^^^^^^^^ 2025-09-12 10:03:36.175126 File "/opt/frigate/frigate/detectors/plugins/openvino.py", line 50, in __init__ 2025-09-12 10:03:36.175133 self.interpreter = self.ov_core.compile_model( 2025-09-12 10:03:36.175140 ^^^^^^^^^^^^^^^^^^^^^^^^^^^ 2025-09-12 10:03:36.175147 File "/usr/local/lib/python3.11/dist-packages/openvino/runtime/ie_api.py", line 543, in compile_model 2025-09-12 10:03:36.175154 super().compile_model(model, device_name, {} if config is None else config), 2025-09-12 10:03:36.175160 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 2025-09-12 10:03:36.175167 RuntimeError: Exception from src/inference/src/cpp/core.cpp:121: 2025-09-12 10:03:36.175174 Exception from src/inference/src/dev/plugin.cpp:58: 2025-09-12 10:03:36.175180 Check 'contexts.count(device_id)' failed at src/plugins/intel_gpu/src/plugin/plugin.cpp:232: 2025-09-12 10:03:36.175187 [GPU] Context was not initialized for 0 device 2025-09-12 10:03:36.175198 2025-09-12 10:03:36.175205 2025-09-12 10:03:36.175211 Added to frigate config.yml in the camera section (didn't like it in the root). ffmpeg: hwaccel_args: preset-vaapi Now gives a bunch of errors, I assume because ffmpeg fails because it can't get the hardware acceleration. Neither system have a /dev/dri although I didn't see the chown error this time. Both lxc .conf files still have the extra /. Did I run the wrong command? I did get different results than the last time I ran this,. One machine was: bash -c "$(curl -fsSL https://raw.githubusercontent.com/remz1337/ProxmoxVED/pr-frigate/ct/frigate.sh)" The other was: bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVED/raw/main/ct/frigate.sh)" I cut and paste from above and I guess I grabbed different lines. Same result though.

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 12, 2025):

@remz1337 You didn't really change the source of build.func , it's still pointing at VED instead of VE.

@Zixim commented on GitHub (Sep 12, 2025): @remz1337 You didn't really change the source of build.func , it's still pointing at VED instead of VE.

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 12, 2025):

@MickLesk do you think apt install ffmpeg is right ?
Asking because this installs ffmpeg version 5.1.7-0+deb12u1, but Jellyfin with iGPU has a totally different version ffmpeg version 7.1.1-Jellyfin.

Interesting find. On my 8.4 system the base OS has 5.1.7 and that seems to be the lastest on Debian Bookworm. On my 9.0.6 system the base is 7.1.1 but the LXC is downloading 5.1.7 when I do an apt-get install ffmpeg.

I hadn't really paid attention but the LXC is Debian 12 instead of Debian 13 so it gets an old ffmpeg.

The script is expecting version 7, the commands from @rommess are actually trying to tell frigate that version 7 exists by pointing to version 5 ffmpeg.

@atrozgrifo commented on GitHub (Sep 12, 2025): > [@MickLesk](https://github.com/MickLesk) do you think `apt install ffmpeg` is right ? > Asking because this installs `ffmpeg version 5.1.7-0+deb12u1`, but Jellyfin with iGPU has a totally different version `ffmpeg version 7.1.1-Jellyfin`. Interesting find. On my 8.4 system the base OS has 5.1.7 and that seems to be the lastest on Debian Bookworm. On my 9.0.6 system the base is 7.1.1 but the LXC is downloading 5.1.7 when I do an apt-get install ffmpeg. I hadn't really paid attention but the LXC is Debian 12 instead of Debian 13 so it gets an old ffmpeg. The script is expecting version 7, the commands from @rommess are actually trying to tell frigate that version 7 exists by pointing to version 5 ffmpeg.

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 12, 2025):

That's really weird, you should not need to install ffmpeg manually. Frigate docker image is built on Debian 12. The script might work on Debian 13, but I haven't tested yet.

I have been testing with my fork and everything installs fine on my end. Could you try that? You will need to go into advanced settings and change some defaults if you want to enable iGPU passthrough:

Change those settings in the advanced settings menu:

priviledged = True
Nvidia passthrough = No

To run the script from my fork:
bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"

@remz1337 commented on GitHub (Sep 12, 2025): That's really weird, you should not need to install ffmpeg manually. Frigate docker image is built on Debian 12. The script might work on Debian 13, but I haven't tested yet. I have been testing with my fork and everything installs fine on my end. Could you try that? You will need to go into advanced settings and change some defaults if you want to enable iGPU passthrough: Change those settings in the advanced settings menu: > priviledged = True > Nvidia passthrough = No To run the script from my fork: `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"`

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 12, 2025):

I have been testing with my fork and everything installs fine on my end. Could you try that?

Fails on Proxmox 9.0x:
✖️ ✖️ This version of Proxmox Virtual Environment is not supported
Requires Proxmox Virtual Environment Version 8.1 or later.

On Promox 8.4:
Seems to be installing a lot more stuff, taking much longer to test.

Done.. Shows test video but then Web UI stops responding. Looking at the frigate log it gets to the line:
frigate.api.fastapi_app INFO : FastAPI started

Then about a minute later it restarts.

"current" log always gets overwritten but I tried doing a tail -f on it but nothing else appears.

If I look at the log in frigate web UI I get one more line:

frigate.watchdog Detection appears to have stopped. Exiting Frigate...

I do have a /dev/dri and it seems to have the right permissions.

ls -ltra /dev/dri
total 0
crw-rw---- 1 root render 226, 128 Sep 10 09:59 renderD128
drwxr-xr-x 3 root video 100 Sep 10 09:59 .
crw-rw---- 1 root video 226, 0 Sep 10 09:59 card0
drw-rw---- 2 root root 80 Sep 10 09:59 by-path
drwxr-xr-x 9 root root 640 Sep 12 11:53 ..

@atrozgrifo commented on GitHub (Sep 12, 2025): > I have been testing with my fork and everything installs fine on my end. Could you try that? Fails on Proxmox 9.0x: ✖️ ✖️ This version of Proxmox Virtual Environment is not supported Requires Proxmox Virtual Environment Version 8.1 or later. On Promox 8.4: Seems to be installing a lot more stuff, taking much longer to test. Done.. Shows test video but then Web UI stops responding. Looking at the frigate log it gets to the line: frigate.api.fastapi_app INFO : FastAPI started Then about a minute later it restarts. "current" log always gets overwritten but I tried doing a tail -f on it but nothing else appears. If I look at the log in frigate web UI I get one more line: frigate.watchdog Detection appears to have stopped. Exiting Frigate... I do have a /dev/dri and it seems to have the right permissions. ls -ltra /dev/dri total 0 crw-rw---- 1 root render 226, 128 Sep 10 09:59 renderD128 drwxr-xr-x 3 root video 100 Sep 10 09:59 . crw-rw---- 1 root video 226, 0 Sep 10 09:59 card0 drw-rw---- 2 root root 80 Sep 10 09:59 by-path drwxr-xr-x 9 root root 640 Sep 12 11:53 ..

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 12, 2025):

@atrozgrifo edit /etc/systemd/system/frigate.service and comment this line:
ExecStartPre=+rm /dev/shm/logs/frigate/current

This should allow to keep the logs. Please paste the logs here once you have them

@remz1337 commented on GitHub (Sep 12, 2025): @atrozgrifo edit `/etc/systemd/system/frigate.service` and comment this line: `ExecStartPre=+rm /dev/shm/logs/frigate/current` This should allow to keep the logs. Please paste the logs here once you have them

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 12, 2025):

This should allow to keep the logs

Nope. Made the change as follows, rebooted the LXC and the logs just get overwritten

[Unit]
Description=Frigate service
After=go2rtc.service
After=create_directories.service
StartLimitIntervalSec=0

[Service]
Type=simple
Restart=always
RestartSec=1
User=root
EnvironmentFile=/etc/frigate.env
# Environment=PLUS_API_KEY=
#ExecStartPre=+rm /dev/shm/logs/frigate/current
ExecStart=/bin/bash -c "bash /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run 2> >(/usr/bin/ts '%%Y-%%m-%%d %%H:%%M:%%.S ' >&2) | /usr/bin/ts '%%Y-%%m-%%d %%H:%%M:%%.S '"
StandardOutput=file:/dev/shm/logs/frigate/current
StandardError=file:/dev/shm/logs/frigate/current

[Install]
WantedBy=multi-user.target

root@frigate16node0:/dev/shm/logs/frigate# cat current
2025-09-12 13:20:48.105574  [INFO] Preparing Frigate...
2025-09-12 13:20:48.357491  [INFO] Starting Frigate...
2025-09-12 13:20:51.288856  [2025-09-12 13:20:51] frigate.util.config            INFO    : Checking if frigate config needs migration...
2025-09-12 13:20:51.297304  [2025-09-12 13:20:51] frigate.util.config            INFO    : frigate config does not need migration...
2025-09-12 13:20:51.466212  [2025-09-12 13:20:51] frigate.util.services          INFO    : Automatically detected vaapi hwaccel for video decoding
2025-09-12 13:20:51.478965  [2025-09-12 13:20:51] frigate.app                    INFO    : Starting Frigate (0.16.1-)
2025-09-12 13:20:51.484866  [2025-09-12 13:20:51] peewee_migrate.logs            INFO    : Starting migrations
2025-09-12 13:20:51.485134  [2025-09-12 13:20:51] peewee_migrate.logs            INFO    : There is nothing to migrate
2025-09-12 13:20:51.499995  [2025-09-12 13:20:51] frigate.app                    INFO    : Recording process started: 1231
2025-09-12 13:20:51.513436  [2025-09-12 13:20:51] frigate.app                    INFO    : Review process started: 1244
2025-09-12 13:20:51.515924  [2025-09-12 13:20:51] frigate.app                    INFO    : go2rtc process pid: 159
2025-09-12 13:20:51.528630  [2025-09-12 13:20:51] detector.detector01            INFO    : Starting detection process: 1256
2025-09-12 13:20:51.544237  [2025-09-12 13:20:51] frigate.app                    INFO    : Output process started: 1272
2025-09-12 13:20:51.580851  [2025-09-12 13:20:51] frigate.app                    INFO    : Camera processor started for test: 1306
2025-09-12 13:20:51.581400  [2025-09-12 13:20:51] frigate.app                    INFO    : Capture process started for test: 1309
2025-09-12 13:20:51.943349  [2025-09-12 13:20:51] frigate.api.fastapi_app        INFO    : Starting FastAPI app
2025-09-12 13:20:52.094097  [2025-09-12 13:20:52] frigate.api.fastapi_app        INFO    : FastAPI started
root@frigate16node0:/dev/shm/logs/frigate#

@atrozgrifo commented on GitHub (Sep 12, 2025): > This should allow to keep the logs Nope. Made the change as follows, rebooted the LXC and the logs just get overwritten ``` [Unit] Description=Frigate service After=go2rtc.service After=create_directories.service StartLimitIntervalSec=0 [Service] Type=simple Restart=always RestartSec=1 User=root EnvironmentFile=/etc/frigate.env # Environment=PLUS_API_KEY= #ExecStartPre=+rm /dev/shm/logs/frigate/current ExecStart=/bin/bash -c "bash /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run 2> >(/usr/bin/ts '%%Y-%%m-%%d %%H:%%M:%%.S ' >&2) | /usr/bin/ts '%%Y-%%m-%%d %%H:%%M:%%.S '" StandardOutput=file:/dev/shm/logs/frigate/current StandardError=file:/dev/shm/logs/frigate/current [Install] WantedBy=multi-user.target ``` ``` root@frigate16node0:/dev/shm/logs/frigate# cat current 2025-09-12 13:20:48.105574 [INFO] Preparing Frigate... 2025-09-12 13:20:48.357491 [INFO] Starting Frigate... 2025-09-12 13:20:51.288856 [2025-09-12 13:20:51] frigate.util.config INFO : Checking if frigate config needs migration... 2025-09-12 13:20:51.297304 [2025-09-12 13:20:51] frigate.util.config INFO : frigate config does not need migration... 2025-09-12 13:20:51.466212 [2025-09-12 13:20:51] frigate.util.services INFO : Automatically detected vaapi hwaccel for video decoding 2025-09-12 13:20:51.478965 [2025-09-12 13:20:51] frigate.app INFO : Starting Frigate (0.16.1-) 2025-09-12 13:20:51.484866 [2025-09-12 13:20:51] peewee_migrate.logs INFO : Starting migrations 2025-09-12 13:20:51.485134 [2025-09-12 13:20:51] peewee_migrate.logs INFO : There is nothing to migrate 2025-09-12 13:20:51.499995 [2025-09-12 13:20:51] frigate.app INFO : Recording process started: 1231 2025-09-12 13:20:51.513436 [2025-09-12 13:20:51] frigate.app INFO : Review process started: 1244 2025-09-12 13:20:51.515924 [2025-09-12 13:20:51] frigate.app INFO : go2rtc process pid: 159 2025-09-12 13:20:51.528630 [2025-09-12 13:20:51] detector.detector01 INFO : Starting detection process: 1256 2025-09-12 13:20:51.544237 [2025-09-12 13:20:51] frigate.app INFO : Output process started: 1272 2025-09-12 13:20:51.580851 [2025-09-12 13:20:51] frigate.app INFO : Camera processor started for test: 1306 2025-09-12 13:20:51.581400 [2025-09-12 13:20:51] frigate.app INFO : Capture process started for test: 1309 2025-09-12 13:20:51.943349 [2025-09-12 13:20:51] frigate.api.fastapi_app INFO : Starting FastAPI app 2025-09-12 13:20:52.094097 [2025-09-12 13:20:52] frigate.api.fastapi_app INFO : FastAPI started root@frigate16node0:/dev/shm/logs/frigate# ```

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 12, 2025):

Then commenting:

Restart=always
RestartSec=1

should do the trick

Everything looks fine from the logs you pasted (hwaccel, detector...). Are you still having issues? What's the resource usage like?

@remz1337 commented on GitHub (Sep 12, 2025): Then commenting: ``` Restart=always RestartSec=1 ``` should do the trick Everything looks fine from the logs you pasted (hwaccel, detector...). Are you still having issues? What's the resource usage like?

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 12, 2025):

Are you still having issues? What's the resource usage like?

It was restarting all the time. Basically non-functional.

I found this in the journal:
Sep 12 13:37:46 frigate16node0 systemd[1]: frigate.service: A process of this unit has been killed by the OOM killer.
Sep 12 13:37:49 frigate16node0 systemd[1]: frigate.service: Failed with result 'oom-kill'.
Sep 12 13:37:49 frigate16node0 systemd[1]: frigate.service: Consumed 5.176s CPU time.

The current log only wrote out once, same contents as before. I assume changing the restart caused frigate to only run once.

Resources now are minimal. 0% Cpu and 69 MiB of RAM. When I first stared the LXC it was about 4% CPU and 670 MiB RAM.

@atrozgrifo commented on GitHub (Sep 12, 2025): > Are you still having issues? What's the resource usage like? It was restarting all the time. Basically non-functional. I found this in the journal: Sep 12 13:37:46 frigate16node0 systemd[1]: frigate.service: A process of this unit has been killed by the OOM killer. Sep 12 13:37:49 frigate16node0 systemd[1]: frigate.service: Failed with result 'oom-kill'. Sep 12 13:37:49 frigate16node0 systemd[1]: frigate.service: Consumed 5.176s CPU time. The current log only wrote out once, same contents as before. I assume changing the restart caused frigate to only run once. Resources now are minimal. 0% Cpu and 69 MiB of RAM. When I first stared the LXC it was about 4% CPU and 670 MiB RAM.

kerem commented

Author

Owner

@rommess commented on GitHub (Sep 12, 2025):

From
bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"

on my DEV proxmox 8.4.13 / 4 x Intel(R) Core(TM) i5-6200U CPU @ 2.30GHz (1 Socket)

install seems good. No more infinite web page loading.

Next step later, add my 3 cameras and then move to the PROD proxmox 8.4.12 / 8 x Intel(R) Core(TM) i3-14100T (1 Socket)

@rommess commented on GitHub (Sep 12, 2025): From `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"` on my DEV proxmox 8.4.13 / 4 x Intel(R) Core(TM) i5-6200U CPU @ 2.30GHz (1 Socket) <img width="759" height="694" alt="Image" src="https://github.com/user-attachments/assets/32f95706-e42e-4232-b7df-9686d61d9cc8" /> install seems good. No more infinite web page loading. Next step later, add my 3 cameras and then move to the PROD proxmox 8.4.12 / 8 x Intel(R) Core(TM) i3-14100T (1 Socket)

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 12, 2025):

@rommess , before adding any cameras, can you enable detection and see if hw acceleration is working fine (both for ffmpeg and object detection)? I would also be interested in seeing inference time and detector CPU utilization with openvino vs onnx using an iGPU

to test D-FINE model with onnx, update your config.yml to this:

detect:
  enabled: true
detectors:
  detector01:
    type: onnx
model:
  model_type: dfine
  width: 320
  height: 320
  input_tensor: nchw
  input_dtype: float
  path: /D-FINE/models/dfine_n_coco.onnx
  labelmap_path: /labelmap/coco-80.txt

or to test YoloX:

detect:
  enabled: true
detectors:
  detector01:
    type: onnx
model:
  model_type: yolox
  width: 416
  height: 416
  input_tensor: nchw
  input_dtype: float_denorm
  path: /models/yolox_tiny.onnx
  labelmap_path: /labelmap/coco-80.txt

both models are already available in the container

@remz1337 commented on GitHub (Sep 12, 2025): @rommess , before adding any cameras, can you enable detection and see if hw acceleration is working fine (both for ffmpeg and object detection)? I would also be interested in seeing inference time and detector CPU utilization with openvino vs onnx using an iGPU to test D-FINE model with onnx, update your config.yml to this: ``` detect: enabled: true detectors: detector01: type: onnx model: model_type: dfine width: 320 height: 320 input_tensor: nchw input_dtype: float path: /D-FINE/models/dfine_n_coco.onnx labelmap_path: /labelmap/coco-80.txt ``` or to test YoloX: ``` detect: enabled: true detectors: detector01: type: onnx model: model_type: yolox width: 416 height: 416 input_tensor: nchw input_dtype: float_denorm path: /models/yolox_tiny.onnx labelmap_path: /labelmap/coco-80.txt ``` both models are already available in the container

kerem commented

Author

Owner

@rommess commented on GitHub (Sep 12, 2025):

hum, enabling detect with the default model make my container no more usable, even in console :(
Did not pay attention, but CPU and RAM are not properly provisionned, so i add 2 CPU + 3 Go of RAM

After that, detect with the openvino model works.

D-FINE model with onnx, inference average 75 ms, CPU 170%

YoloX, similare to D-FINE

Remember, this hardware is an old laptop (x270 for DEV purpose) and should perform better on M70q or similar

@rommess commented on GitHub (Sep 12, 2025): hum, enabling detect with the default model make my container no more usable, even in console :( Did not pay attention, but CPU and RAM are not properly provisionned, so i add 2 CPU + 3 Go of RAM <img width="1469" height="699" alt="Image" src="https://github.com/user-attachments/assets/2a91fefa-2c3c-4d2a-aed5-afa7d7957617" /> After that, detect with the openvino model works. <img width="1844" height="517" alt="Image" src="https://github.com/user-attachments/assets/1da5c8cc-a0e1-4748-a220-79d527c1b6ac" /> D-FINE model with onnx, inference average 75 ms, CPU 170% <img width="1849" height="527" alt="Image" src="https://github.com/user-attachments/assets/ebd90aca-6ddb-402c-ba81-629abe6d85e0" /> YoloX, similare to D-FINE <img width="1812" height="197" alt="Image" src="https://github.com/user-attachments/assets/17038186-0769-4de1-b66c-1a8548ba1113" /> Remember, this hardware is an old laptop (x270 for DEV purpose) and should perform better on M70q or similar

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 13, 2025):

Thanks @rommess, I'll increase the default CPU and RAM. Can you paste the frigate logs after running D-FINE for a few minutes? I was expecting higher iGPU usage, so maybe something is wrong

@remz1337 commented on GitHub (Sep 13, 2025): Thanks @rommess, I'll increase the default CPU and RAM. Can you paste the frigate logs after running D-FINE for a few minutes? I was expecting higher iGPU usage, so maybe something is wrong

kerem commented

Author

Owner

@gordyb83 commented on GitHub (Sep 13, 2025):

anyone have an idea what keeps failing on my attempts to get this going its driving me nuts

✔️ Downloaded Frigate source
⠦ Building Nginx with Custom Modules
[ERROR] in line 224: exit code 0: while executing command "$@" > /dev/null 2>&1
/dev/fd/63: line 74: pop_var_context: head of shell_variables not a function context
/dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context

@gordyb83 commented on GitHub (Sep 13, 2025): anyone have an idea what keeps failing on my attempts to get this going its driving me nuts ✔️ Downloaded Frigate source ⠦ Building Nginx with Custom Modules [ERROR] in line 224: exit code 0: while executing command "$@" > /dev/null 2>&1 /dev/fd/63: line 74: pop_var_context: head of shell_variables not a function context /dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context

kerem commented

Author

Owner

@rommess commented on GitHub (Sep 13, 2025):

@remz1337 Here we go

logs:
frigate-logs-dfine_n_coco.onnx.txt

doesn't seems to use iGPU with D-FINE model

@rommess commented on GitHub (Sep 13, 2025): @remz1337 Here we go logs: [frigate-logs-dfine_n_coco.onnx.txt](https://github.com/user-attachments/files/22311862/frigate-logs-dfine_n_coco.onnx.txt) doesn't seems to use iGPU with D-FINE model <img width="1837" height="539" alt="Image" src="https://github.com/user-attachments/assets/a706f498-ca16-4a71-b74f-fb76d8974242" />

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 14, 2025):

@hawkeye217 and @NickM-27 sorry to bother you here, could you help us troubleshoot why onnx isn't leveraging iGPU?
The only thing I see in the logs is:
[W:onnxruntime:, graph.cc:109 MergeShapeInfo] Error merging shape info for output. 'linear_43' source:{1200,1} target:{1200}. Falling back to lenient merge.
but I don't think this should be what's preventing leveraging the iGPU. Any idea as to where we should look?

@remz1337 commented on GitHub (Sep 14, 2025): @hawkeye217 and @NickM-27 sorry to bother you here, could you help us troubleshoot why onnx isn't leveraging iGPU? The only thing I see in the logs is: `[W:onnxruntime:, graph.cc:109 MergeShapeInfo] Error merging shape info for output. 'linear_43' source:{1200,1} target:{1200}. Falling back to lenient merge. ` but I don't think this should be what's preventing leveraging the iGPU. Any idea as to where we should look?

kerem commented

Author

Owner

@NickM-27 commented on GitHub (Sep 14, 2025):

When you use Intel it is recommended to use the openvino detector, not onnx detector. There is some issue in onnxruntime that causes it to fail when trying to use openvino, and openvino is more efficient anyway

@NickM-27 commented on GitHub (Sep 14, 2025): When you use Intel it is recommended to use the openvino detector, not onnx detector. There is some issue in onnxruntime that causes it to fail when trying to use openvino, and openvino is more efficient anyway

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 14, 2025):

    ______     _             __     
   / ____/____(_)___ _____ _/ /____ 
  / /_  / ___/ / __ `/ __ `/ __/ _ \
 / __/ / /  / / /_/ / /_/ / /_/  __/
/_/   /_/  /_/\__, /\__,_/\__/\___/ 
             /____/                 
  ✖️     ✖️  This version of Proxmox Virtual Environment is not supported
Requires Proxmox Virtual Environment Version 8.1 or later.
Exiting...
root@pve:~# pveversion
pve-manager/9.0.9/117b893e0e6a4fee (running kernel: 6.14.11-2-pve)
root@pve:~#

🤷

@Zixim commented on GitHub (Sep 14, 2025): ``` ______ _ __ / ____/____(_)___ _____ _/ /____ / /_ / ___/ / __ `/ __ `/ __/ _ \ / __/ / / / / /_/ / /_/ / /_/ __/ /_/ /_/ /_/\__, /\__,_/\__/\___/ /____/ ✖️ ✖️ This version of Proxmox Virtual Environment is not supported Requires Proxmox Virtual Environment Version 8.1 or later. Exiting... root@pve:~# pveversion pve-manager/9.0.9/117b893e0e6a4fee (running kernel: 6.14.11-2-pve) root@pve:~# ``` 🤷

kerem commented

github.com/remz1337/ProxmoxVE@b1211b9fbb/misc/build.func (L70-L78)

Author

Owner

@Lastof commented on GitHub (Sep 14, 2025):

    ______     _             __     
   / ____/____(_)___ _____ _/ /____ 
  / /_  / ___/ / __ `/ __ `/ __/ _ \
 / __/ / /  / / /_/ / /_/ / /_/  __/
/_/   /_/  /_/\__, /\__,_/\__/\___/ 
             /____/                 
  ✖️     ✖️  This version of Proxmox Virtual Environment is not supported
Requires Proxmox Virtual Environment Version 8.1 or later.
Exiting...
root@pve:~# pveversion
pve-manager/9.0.9/117b893e0e6a4fee (running kernel: 6.14.11-2-pve)
root@pve:~#

🤷

Looks like the version of /misc/build.func in @remz1337's ProxmoxVE repo still uses a version of pve_check that only checks for pveversion starting with 8, which means it fails on PVE9

Their version of ProxmoxVED uses the newer version, so just needs that version of build.func merging accross (assuming that won't break other things).

@Lastof commented on GitHub (Sep 14, 2025): > ``` > ______ _ __ > / ____/____(_)___ _____ _/ /____ > / /_ / ___/ / __ `/ __ `/ __/ _ \ > / __/ / / / / /_/ / /_/ / /_/ __/ > /_/ /_/ /_/\__, /\__,_/\__/\___/ > /____/ > ✖️ ✖️ This version of Proxmox Virtual Environment is not supported > Requires Proxmox Virtual Environment Version 8.1 or later. > Exiting... > root@pve:~# pveversion > pve-manager/9.0.9/117b893e0e6a4fee (running kernel: 6.14.11-2-pve) > root@pve:~# > ``` > > 🤷 Looks like the version of /misc/build.func in @remz1337's ProxmoxVE repo still uses a version of pve_check that only checks for pveversion starting with 8, which means it fails on PVE9 https://github.com/remz1337/ProxmoxVE/blob/b1211b9fbb630526e59b4c9ec8e41d4d5ab5ab1e/misc/build.func#L70-L78 Their version of ProxmoxVED uses the newer version, so just needs that version of build.func merging accross (assuming that won't break other things).

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 15, 2025):

I've disabled PVE version check, should work with 9.x now.

@remz1337 commented on GitHub (Sep 15, 2025): I've disabled PVE version check, should work with 9.x now.

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 16, 2025):

@remz1337 , still doesn't install on a PVE host with iGPU :
installed with : bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"

  ⚙️  Using Default Settings on node pve (🔍  Verbose)
  🖥️  Operating System: debian
  🌟  Version: 12
  📦  Container Type: Unprivileged
  💾  Disk Size: 50 GB
  🧠  CPU Cores: 4
  🛠️  RAM Size: 4096 MiB
  🆔  Container ID: 100
Add common sudo user with SSH access: no
Enable Shared Mount: no
Configure Postfix Satellite: no
Passthrough Nvidia GPU: yes
  🔍  Verbose Mode: Enabled
  🚀  Creating a Frigate LXC using the above default settings
  
  ✔️   Using iso for Template Storage.
  ✔️   Using lxc-zfs for Container Storage.
  ✔️   LXC Template List Updated
  ✔️   LXC Template 'debian-12-standard_12.12-1_amd64.tar.zst' is ready to use.
  ✔️   LXC Container 100 was successfully created.
  ✔️   Started LXC Container
  ✔️   Customized LXC Container
Customizing LXC creation
  ✔️   Rebooted LXC
  ✔️   Set up Container OS
  ✔️   Network Connected: 192.168.11.198 
  ✔️   IPv4 Internet Connected
  ✖️   IPv6 Internet Not Connected
  ✔️   GitHub DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
Hit:1 http://deb.debian.org/debian bookworm InRelease
Hit:2 http://security.debian.org bookworm-security InRelease
Hit:3 http://deb.debian.org/debian bookworm-updates InRelease
Reading package lists... Done
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
Calculating upgrade... Done
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
  ✔️   Updated Container OS
  ✔️   Set up Container
  ✔️   Installed APT proxy client
  ✔️   Installed sudo
NVIDIA drivers not found!
⚠  User exited script 

root@pve:~# pveversion
pve-manager/9.0.9/117b893e0e6a4fee (running kernel: 6.14.11-2-pve)
root@pve:~#

@Zixim commented on GitHub (Sep 16, 2025): @remz1337 , still doesn't install on a PVE host with iGPU : installed with : `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"` ``` ⚙️ Using Default Settings on node pve (🔍 Verbose) 🖥️ Operating System: debian 🌟 Version: 12 📦 Container Type: Unprivileged 💾 Disk Size: 50 GB 🧠 CPU Cores: 4 🛠️ RAM Size: 4096 MiB 🆔 Container ID: 100 Add common sudo user with SSH access: no Enable Shared Mount: no Configure Postfix Satellite: no Passthrough Nvidia GPU: yes 🔍 Verbose Mode: Enabled 🚀 Creating a Frigate LXC using the above default settings ✔️ Using iso for Template Storage. ✔️ Using lxc-zfs for Container Storage. ✔️ LXC Template List Updated ✔️ LXC Template 'debian-12-standard_12.12-1_amd64.tar.zst' is ready to use. ✔️ LXC Container 100 was successfully created. ✔️ Started LXC Container ✔️ Customized LXC Container Customizing LXC creation ✔️ Rebooted LXC ✔️ Set up Container OS ✔️ Network Connected: 192.168.11.198 ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ GitHub DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) Hit:1 http://deb.debian.org/debian bookworm InRelease Hit:2 http://security.debian.org bookworm-security InRelease Hit:3 http://deb.debian.org/debian bookworm-updates InRelease Reading package lists... Done Reading package lists... Done Building dependency tree... Done Reading state information... Done Calculating upgrade... Done 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. ✔️ Updated Container OS ✔️ Set up Container ✔️ Installed APT proxy client ✔️ Installed sudo NVIDIA drivers not found! ⚠ User exited script root@pve:~# pveversion pve-manager/9.0.9/117b893e0e6a4fee (running kernel: 6.14.11-2-pve) root@pve:~# ```

kerem commented

Author

Owner

@mattosaurus commented on GitHub (Sep 16, 2025):

@remz1337 , still doesn't install on a PVE host with iGPU : installed with : bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"

  ⚙️  Using Default Settings on node pve (🔍  Verbose)
  🖥️  Operating System: debian
  🌟  Version: 12
  📦  Container Type: Unprivileged
  💾  Disk Size: 50 GB
  🧠  CPU Cores: 4
  🛠️  RAM Size: 4096 MiB
  🆔  Container ID: 100
Add common sudo user with SSH access: no
Enable Shared Mount: no
Configure Postfix Satellite: no
Passthrough Nvidia GPU: yes
  🔍  Verbose Mode: Enabled
  🚀  Creating a Frigate LXC using the above default settings
  
  ✔️   Using iso for Template Storage.
  ✔️   Using lxc-zfs for Container Storage.
  ✔️   LXC Template List Updated
  ✔️   LXC Template 'debian-12-standard_12.12-1_amd64.tar.zst' is ready to use.
  ✔️   LXC Container 100 was successfully created.
  ✔️   Started LXC Container
  ✔️   Customized LXC Container
Customizing LXC creation
  ✔️   Rebooted LXC
  ✔️   Set up Container OS
  ✔️   Network Connected: 192.168.11.198 
  ✔️   IPv4 Internet Connected
  ✖️   IPv6 Internet Not Connected
  ✔️   GitHub DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
Hit:1 http://deb.debian.org/debian bookworm InRelease
Hit:2 http://security.debian.org bookworm-security InRelease
Hit:3 http://deb.debian.org/debian bookworm-updates InRelease
Reading package lists... Done
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
Calculating upgrade... Done
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
  ✔️   Updated Container OS
  ✔️   Set up Container
  ✔️   Installed APT proxy client
  ✔️   Installed sudo
NVIDIA drivers not found!
⚠  User exited script 

root@pve:~# pveversion
pve-manager/9.0.9/117b893e0e6a4fee (running kernel: 6.14.11-2-pve)
root@pve:~#

If you're using an iGPU rather than a Nvidia card you need to use the advanced installation and select no to Nvidia.

@mattosaurus commented on GitHub (Sep 16, 2025): > [@remz1337](https://github.com/remz1337) , still doesn't install on a PVE host with iGPU : installed with : `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"` > > ``` > ⚙️ Using Default Settings on node pve (🔍 Verbose) > 🖥️ Operating System: debian > 🌟 Version: 12 > 📦 Container Type: Unprivileged > 💾 Disk Size: 50 GB > 🧠 CPU Cores: 4 > 🛠️ RAM Size: 4096 MiB > 🆔 Container ID: 100 > Add common sudo user with SSH access: no > Enable Shared Mount: no > Configure Postfix Satellite: no > Passthrough Nvidia GPU: yes > 🔍 Verbose Mode: Enabled > 🚀 Creating a Frigate LXC using the above default settings > > ✔️ Using iso for Template Storage. > ✔️ Using lxc-zfs for Container Storage. > ✔️ LXC Template List Updated > ✔️ LXC Template 'debian-12-standard_12.12-1_amd64.tar.zst' is ready to use. > ✔️ LXC Container 100 was successfully created. > ✔️ Started LXC Container > ✔️ Customized LXC Container > Customizing LXC creation > ✔️ Rebooted LXC > ✔️ Set up Container OS > ✔️ Network Connected: 192.168.11.198 > ✔️ IPv4 Internet Connected > ✖️ IPv6 Internet Not Connected > ✔️ GitHub DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) > Hit:1 http://deb.debian.org/debian bookworm InRelease > Hit:2 http://security.debian.org bookworm-security InRelease > Hit:3 http://deb.debian.org/debian bookworm-updates InRelease > Reading package lists... Done > Reading package lists... Done > Building dependency tree... Done > Reading state information... Done > Calculating upgrade... Done > 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. > ✔️ Updated Container OS > ✔️ Set up Container > ✔️ Installed APT proxy client > ✔️ Installed sudo > NVIDIA drivers not found! > ⚠ User exited script > > root@pve:~# pveversion > pve-manager/9.0.9/117b893e0e6a4fee (running kernel: 6.14.11-2-pve) > root@pve:~# > ``` If you're using an iGPU rather than a Nvidia card you need to use the advanced installation and select no to Nvidia.

kerem commented

Author

Owner

@mattosaurus commented on GitHub (Sep 16, 2025):

The script works for me other than an issue building one of the models.

Running on a Dell OptiPlex 7050 with an iGPU and USB Coral.

  🧩  Using Advanced Settings on node pve
  🖥️  Operating System: debian
  🌟  Version: 12
  📦  Container Type: Privileged
  🔐  Root Password: ********
  🆔  Container ID: 101
  🏠  Hostname: frigate
  💾  Disk Size: 30 GB
  🧠  CPU Cores: 4
  🛠️  RAM Size: 4096 MiB
  🌉  Bridge: vmbr0
  📡  IP Address: dhcp
  🌐  Gateway IP Address: Default
  📡  APT-Cacher IP Address: Default
  🚫  Disable IPv6: no
  ⚙️  Interface MTU Size: Default
  🔍  DNS Search Domain: Host
  📡  DNS Server IP Address: Host
Enable Firewall: 1
  🏷️  Vlan: Default
  📡  Tags: nvr
  🔑  Root SSH Access: yes
  🗂️  Enable FUSE Support: no
Add common sudo user with SSH access: no
Enable Shared Mount: no
Configure Postfix Satellite: no
Nvidia GPU Passthrough: no
  🔍  Verbose Mode: no
  🚀  Creating a Frigate LXC using the above advanced settings
  ✔️   Using local for Template Storage.
  ✔️   Using local-lvm for Container Storage.
  ✔️   LXC Template 'debian-12-standard_12.12-1_amd64.tar.zst' is ready to use.
  ✔️   LXC Container 101 was successfully created.
  ✔️   Started LXC Container
  ✔️   Customized LXC Container
Customizing LXC creation
  ✔️   Rebooted LXC
  ✔️   Set up Container OS
  ✔️   Network Connected: 192.168.8.217 fdd8:6628:b06b:c747:be24:11ff:fe0c:2a79 
  ✔️   IPv4 Internet Connected
  ✖️   IPv6 Internet Not Connected
  ✔️   GitHub DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
  ✔️   Updated Container OS
  ✔️   Set up Container
  ✔️   Installed APT proxy client
  ✔️   Installed sudo
  ✔️   Post install script completed.
  ✔️   Set up Container OS
  ✔️   Network Connected: 192.168.8.217 fdd8:6628:b06b:c747:be24:11ff:fe0c:2a79 
  ✔️   IPv4 Internet Connected
  ✖️   IPv6 Internet Not Connected
  ✔️   GitHub DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
  ✔️   Updated Container OS
  ✔️   Installed Dependencies
  ✔️   Set Up Hardware Acceleration
  ✔️   Setup environment
  ✔️   Downloaded Frigate source
  ✔️   Built Nginx
  ✔️   Built SQLite
  ✔️   Installed go2rtc
  ✔️   Installed Tempio
  ✔️   Built libUSB
  ✔️   Installed Pip
  ✔️   Installed Frigate Dependencies
  ✔️   Built pysqlite3
  ✔️   Installed NodeJS
  ✔️   Downloaded Coral TPU Model
  ✔️   Downloaded CPU Model
  ✔️   Built Audio Models
  ✔️   Built HailoRT
  ✔️   Installed OpenVino Runtime and Dev library
  ✔️   Downloaded OpenVino Model
  ⠦   Building D-FINE Model/dev/fd/63: line 223: 42628 Segmentation fault      "$@" > /dev/null 2>&1
  ⠇   Building D-FINE Model
[ERROR] in line 224: exit code 0: while executing command "$@" > /dev/null 2>&1

[ERROR] in line 227: exit code 0: while executing command "$@" > /dev/null 2>&1
  ✔️   Built D-FINE Model
  ✔️   Downloaded YoloX Model
  ✔️   Installed Frigate
  ✔️   AVX or SSE 4.2 Support Detected
  ✔️   Configured Openvino Object Detection Model
  ✔️   Created Services
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
  ✔️   Cleaned
Don't forget to edit the Frigate config file (/config/config.yml) and reboot. Example configuration at https://docs.frigate.video/configuration/
  ✔️   Set Container to Normal Resources
  ✔️   Completed Successfully!

  🚀  Frigate setup has been successfully initialized!
  💡   Access it using the following URL:
    🌐  http://192.168.8.217:5000

Interestingly the memory and cores don't seem to be getting applied as the LXC has been created with 2GiB RAM and 2 cores rather than the expected defaults.

@mattosaurus commented on GitHub (Sep 16, 2025): The script works for me other than an issue building one of the models. Running on a Dell OptiPlex 7050 with an iGPU and USB Coral. ``` 🧩 Using Advanced Settings on node pve 🖥️ Operating System: debian 🌟 Version: 12 📦 Container Type: Privileged 🔐 Root Password: ******** 🆔 Container ID: 101 🏠 Hostname: frigate 💾 Disk Size: 30 GB 🧠 CPU Cores: 4 🛠️ RAM Size: 4096 MiB 🌉 Bridge: vmbr0 📡 IP Address: dhcp 🌐 Gateway IP Address: Default 📡 APT-Cacher IP Address: Default 🚫 Disable IPv6: no ⚙️ Interface MTU Size: Default 🔍 DNS Search Domain: Host 📡 DNS Server IP Address: Host Enable Firewall: 1 🏷️ Vlan: Default 📡 Tags: nvr 🔑 Root SSH Access: yes 🗂️ Enable FUSE Support: no Add common sudo user with SSH access: no Enable Shared Mount: no Configure Postfix Satellite: no Nvidia GPU Passthrough: no 🔍 Verbose Mode: no 🚀 Creating a Frigate LXC using the above advanced settings ✔️ Using local for Template Storage. ✔️ Using local-lvm for Container Storage. ✔️ LXC Template 'debian-12-standard_12.12-1_amd64.tar.zst' is ready to use. ✔️ LXC Container 101 was successfully created. ✔️ Started LXC Container ✔️ Customized LXC Container Customizing LXC creation ✔️ Rebooted LXC ✔️ Set up Container OS ✔️ Network Connected: 192.168.8.217 fdd8:6628:b06b:c747:be24:11ff:fe0c:2a79 ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ GitHub DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) ✔️ Updated Container OS ✔️ Set up Container ✔️ Installed APT proxy client ✔️ Installed sudo ✔️ Post install script completed. ✔️ Set up Container OS ✔️ Network Connected: 192.168.8.217 fdd8:6628:b06b:c747:be24:11ff:fe0c:2a79 ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ GitHub DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) ✔️ Updated Container OS ✔️ Installed Dependencies ✔️ Set Up Hardware Acceleration ✔️ Setup environment ✔️ Downloaded Frigate source ✔️ Built Nginx ✔️ Built SQLite ✔️ Installed go2rtc ✔️ Installed Tempio ✔️ Built libUSB ✔️ Installed Pip ✔️ Installed Frigate Dependencies ✔️ Built pysqlite3 ✔️ Installed NodeJS ✔️ Downloaded Coral TPU Model ✔️ Downloaded CPU Model ✔️ Built Audio Models ✔️ Built HailoRT ✔️ Installed OpenVino Runtime and Dev library ✔️ Downloaded OpenVino Model ⠦ Building D-FINE Model/dev/fd/63: line 223: 42628 Segmentation fault "$@" > /dev/null 2>&1 ⠇ Building D-FINE Model [ERROR] in line 224: exit code 0: while executing command "$@" > /dev/null 2>&1 [ERROR] in line 227: exit code 0: while executing command "$@" > /dev/null 2>&1 ✔️ Built D-FINE Model ✔️ Downloaded YoloX Model ✔️ Installed Frigate ✔️ AVX or SSE 4.2 Support Detected ✔️ Configured Openvino Object Detection Model ✔️ Created Services Reading package lists... Done Building dependency tree... Done Reading state information... Done 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. Reading package lists... Done Building dependency tree... Done Reading state information... Done ✔️ Cleaned Don't forget to edit the Frigate config file (/config/config.yml) and reboot. Example configuration at https://docs.frigate.video/configuration/ ✔️ Set Container to Normal Resources ✔️ Completed Successfully! 🚀 Frigate setup has been successfully initialized! 💡 Access it using the following URL: 🌐 http://192.168.8.217:5000 ``` Interestingly the memory and cores don't seem to be getting applied as the LXC has been created with 2GiB RAM and 2 cores rather than the expected defaults.

kerem commented

Author

Owner

@Adrien40 commented on GitHub (Sep 16, 2025):

Hello,
installed with : bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"
Intel N150 (iGPU) - Proxmox 9.0.9
Advanced installation and select no to Nvidia, i left the rest as default.

🧩 Using Advanced Settings on node proxmox
🖥️ Operating System: debian
🌟 Version: 12
📦 Container Type: Unprivileged
🔐 Root Password: Automatic Login
🆔 Container ID: 103
🏠 Hostname: frigate
💾 Disk Size: 50 GB
🧠 CPU Cores: 4
🛠️ RAM Size: 4096 MiB
🌉 Bridge: vmbr0
📡 IP Address: dhcp
🌐 Gateway IP Address: Default
📡 APT-Cacher IP Address: Default
🚫 Disable IPv6: no
⚙️ Interface MTU Size: Default
🔍 DNS Search Domain: Host
📡 DNS Server IP Address: Host
Enable Firewall: 1
🏷️ Vlan: Default
📡 Tags: remz1337;nvr
🔑 Root SSH Access: no
🗂️ Enable FUSE Support: no
Add common sudo user with SSH access: no
Enable Shared Mount: no
Configure Postfix Satellite: no
Nvidia GPU Passthrough: no
🔍 Verbose Mode: no
🚀 Creating a Frigate LXC using the above advanced settings
✔️ Using local for Template Storage.
✔️ Using local-lvm for Container Storage.
✔️ LXC Template List Updated
✔️ LXC Template 'debian-12-standard_12.12-1_amd64.tar.zst' is ready to use.
✔️ LXC Container 103 was successfully created.
✔️ Started LXC Container
✔️ Customized LXC Container
Customizing LXC creation
✔️ Rebooted LXC
✔️ Set up Container OS
✔️ Network Connected: 192.168.1.113 2a01:e0a:c36:e660:be24:11ff:fe40:178a
✔️ IPv4 Internet Connected
✔️ IPv6 Internet Connected
✔️ GitHub DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
✔️ Updated Container OS
✔️ Set up Container
✔️ Installed APT proxy client
✔️ Installed sudo
✔️ Post install script completed.
✔️ Set up Container OS
✔️ Network Connected: 192.168.1.113 2a01:e0a:c36:e660:be24:11ff:fe40:178a
✔️ IPv4 Internet Connected
✔️ IPv6 Internet Connected
✔️ GitHub DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
✔️ Updated Container OS
✔️ Installed Dependencies
✔️ Set Up Hardware Acceleration
✔️ Setup environment
✔️ Downloaded Frigate source
✔️ Built Nginx
✔️ Built SQLite
✔️ Installed go2rtc
✔️ Installed Tempio
✔️ Built libUSB
✔️ Installed Pip
✔️ Installed Frigate Dependencies
✔️ Built pysqlite3
✔️ Installed NodeJS
✔️ Downloaded Coral TPU Model
✔️ Downloaded CPU Model
✔️ Built Audio Models
✔️ Built HailoRT
✔️ Installed OpenVino Runtime and Dev library
✔️ Downloaded OpenVino Model
⠇ Building D-FINE Model/dev/fd/63: line 223: 39763 Segmentation fault "$@" > /dev/null 2>&1
⠙ Building D-FINE Model
[ERROR] in line 224: exit code 0: while executing command "$@" > /dev/null 2>&1

[ERROR] in line 227: exit code 0: while executing command "$@" > /dev/null 2>&1
✔️ Built D-FINE Model
✔️ Downloaded YoloX Model
✔️ Installed Frigate
✔️ AVX or SSE 4.2 Support Detected
✔️ Configured Openvino Object Detection Model
✔️ Created Services
✔️ Customized Container
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
✔️ Cleaned
Don't forget to edit the Frigate config file (/config/config.yml) and reboot. Example configuration at https://docs.frigate.video/configuration/
✔️ Set Container to Normal Resources
✔️ Completed Successfully!

🚀 Frigate setup has been successfully initialized!
💡 Access it using the following URL:
🌐 http://192.168.1.113:5000

@Adrien40 commented on GitHub (Sep 16, 2025): Hello, installed with : `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"` Intel N150 (iGPU) - Proxmox 9.0.9 Advanced installation and select no to Nvidia, i left the rest as default. 🧩 Using Advanced Settings on node proxmox 🖥️ Operating System: debian 🌟 Version: 12 📦 Container Type: Unprivileged 🔐 Root Password: Automatic Login 🆔 Container ID: 103 🏠 Hostname: frigate 💾 Disk Size: 50 GB 🧠 CPU Cores: 4 🛠️ RAM Size: 4096 MiB 🌉 Bridge: vmbr0 📡 IP Address: dhcp 🌐 Gateway IP Address: Default 📡 APT-Cacher IP Address: Default 🚫 Disable IPv6: no ⚙️ Interface MTU Size: Default 🔍 DNS Search Domain: Host 📡 DNS Server IP Address: Host Enable Firewall: 1 🏷️ Vlan: Default 📡 Tags: remz1337;nvr 🔑 Root SSH Access: no 🗂️ Enable FUSE Support: no Add common sudo user with SSH access: no Enable Shared Mount: no Configure Postfix Satellite: no Nvidia GPU Passthrough: no 🔍 Verbose Mode: no 🚀 Creating a Frigate LXC using the above advanced settings ✔️ Using local for Template Storage. ✔️ Using local-lvm for Container Storage. ✔️ LXC Template List Updated ✔️ LXC Template 'debian-12-standard_12.12-1_amd64.tar.zst' is ready to use. ✔️ LXC Container 103 was successfully created. ✔️ Started LXC Container ✔️ Customized LXC Container Customizing LXC creation ✔️ Rebooted LXC ✔️ Set up Container OS ✔️ Network Connected: 192.168.1.113 2a01:e0a:c36:e660:be24:11ff:fe40:178a ✔️ IPv4 Internet Connected ✔️ IPv6 Internet Connected ✔️ GitHub DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) ✔️ Updated Container OS ✔️ Set up Container ✔️ Installed APT proxy client ✔️ Installed sudo ✔️ Post install script completed. ✔️ Set up Container OS ✔️ Network Connected: 192.168.1.113 2a01:e0a:c36:e660:be24:11ff:fe40:178a ✔️ IPv4 Internet Connected ✔️ IPv6 Internet Connected ✔️ GitHub DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) ✔️ Updated Container OS ✔️ Installed Dependencies ✔️ Set Up Hardware Acceleration ✔️ Setup environment ✔️ Downloaded Frigate source ✔️ Built Nginx ✔️ Built SQLite ✔️ Installed go2rtc ✔️ Installed Tempio ✔️ Built libUSB ✔️ Installed Pip ✔️ Installed Frigate Dependencies ✔️ Built pysqlite3 ✔️ Installed NodeJS ✔️ Downloaded Coral TPU Model ✔️ Downloaded CPU Model ✔️ Built Audio Models ✔️ Built HailoRT ✔️ Installed OpenVino Runtime and Dev library ✔️ Downloaded OpenVino Model ⠇ Building D-FINE Model/dev/fd/63: line 223: 39763 Segmentation fault "$@" > /dev/null 2>&1 ⠙ Building D-FINE Model [ERROR] in line 224: exit code 0: while executing command "$@" > /dev/null 2>&1 [ERROR] in line 227: exit code 0: while executing command "$@" > /dev/null 2>&1 ✔️ Built D-FINE Model ✔️ Downloaded YoloX Model ✔️ Installed Frigate ✔️ AVX or SSE 4.2 Support Detected ✔️ Configured Openvino Object Detection Model ✔️ Created Services ✔️ Customized Container Reading package lists... Done Building dependency tree... Done Reading state information... Done 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. Reading package lists... Done Building dependency tree... Done Reading state information... Done ✔️ Cleaned Don't forget to edit the Frigate config file (/config/config.yml) and reboot. Example configuration at https://docs.frigate.video/configuration/ ✔️ Set Container to Normal Resources ✔️ Completed Successfully! 🚀 Frigate setup has been successfully initialized! 💡 Access it using the following URL: 🌐 http://192.168.1.113:5000

kerem commented

Author

Owner

@vibr77 commented on GitHub (Sep 16, 2025):

I was able to install frigate but I am getting wss api error 500, same with Safari / Google, where are the log file ?

@vibr77 commented on GitHub (Sep 16, 2025): I was able to install frigate but I am getting wss api error 500, same with Safari / Google, where are the log file ?

kerem commented

Author

Owner

@mattosaurus commented on GitHub (Sep 16, 2025):

tail -n 50 /dev/shm/logs/frigate/current

@mattosaurus commented on GitHub (Sep 16, 2025): `tail -n 50 /dev/shm/logs/frigate/current`

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 16, 2025):

@vibr77

root@frigate:~# ls -l /dev/shm/logs
total 0
drwxrwxrwx 2 root root 60 Sep 16 11:06 frigate
drwxrwxrwx 2 root root 60 Sep 16 11:06 go2rtc
drwxrwxrwx 2 root root 60 Sep 16 11:06 nginx

@Zixim commented on GitHub (Sep 16, 2025): @vibr77 ``` root@frigate:~# ls -l /dev/shm/logs total 0 drwxrwxrwx 2 root root 60 Sep 16 11:06 frigate drwxrwxrwx 2 root root 60 Sep 16 11:06 go2rtc drwxrwxrwx 2 root root 60 Sep 16 11:06 nginx ```

kerem commented

Author

Owner

@mattosaurus commented on GitHub (Sep 16, 2025):

The edgetpu_model.tflite file seems to be missing.

2025-09-16 09:46:20.979452 ValueError: Could not open '/edgetpu_model.tflite'.

Th install script has ✔️ Downloaded Coral TPU Model which I'm assuming is the required file. Is the path for this different?

@mattosaurus commented on GitHub (Sep 16, 2025): The edgetpu_model.tflite file seems to be missing. 2025-09-16 09:46:20.979452 ValueError: Could not open '/edgetpu_model.tflite'. Th install script has `✔️ Downloaded Coral TPU Model` which I'm assuming is the required file. Is the path for this different?

kerem commented

Author

Owner

@vibr77 commented on GitHub (Sep 16, 2025):

Hello,
I am getting this error...

2025-09-16 11:38:09.388144 192.168.4.29 - - [16/Sep/2025:11:38:09 +0200] "GET /fonts/Inter-Medium.woff2 HTTP/1.1" 200 111380 "http://192.168.3.35:5000/" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.0 Safari/605.1.15" "-"
2025-09-16 11:38:09.602684 2025/09/16 11:38:09 [error] 229#229: *1431 connect() failed (111: Connection refused) while connecting to upstream, client: 192.168.4.29, server: , request: "GET /api/profile HTTP/1.1", subrequest: "/auth", upstream: "http://127.0.0.1:5001/auth", host: "192.168.3.35:5000", referrer: "http://192.168.3.35:5000/"
2025-09-16 11:38:09.602751 2025/09/16 11:38:09 [error] 229#229: *1431 auth request unexpected status: 502 while sending to client, client: 192.168.4.29, server: , request: "GET /api/profile HTTP/1.1", host: "192.168.3.35:5000", referrer: "http://192.168.3.35:5000/"
2025-09-16 11:38:09.602759 192.168.4.29 - - [16/Sep/2025:11:38:09 +0200] "GET /api/profile HTTP/1.1" 500 177 "http://192.168.3.35:5000/" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.0 Safari/605.1.15" "-"
2025-09-16 11:38:09.602765 2025/09/16 11:38:09 [error] 229#229: *1433 connect() failed (111: Connection refused) while connecting to upstream, client: 192.168.4.29, server: , request: "GET /api/profile HTTP/1.1", subrequest: "/auth", upstream: "http://127.0.0.1:5001/auth", host: "192.168.3.35:5000", referrer: "http://192.168.3.35:5000/"
2025-09-16 11:38:09.602771 2025/09/16 11:38:09 [error] 229#229: *1433 auth request unexpected status: 502 while sending to client, client: 192.168.4.29, server: , request: "GET /api/profile HTTP/1.1", host: "192.168.3.35:5000", referrer: "http://192.168.3.35:5000/"
2025-09-16 11:38:09.602777 192.168.4.29 - - [16/Sep/2025:11:38:09 +0200] "GET /api/profile HTTP/1.1" 500 177 "http://192.168.3.35:5000/" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.0 Safari/605.1.15" "-"
2025-09-16 11:38:09.602783 2025/09/16 11:38:09 [error] 229#229: *1432 connect() failed (111: Connection refused) while connecting to upstream, client: 192.168.4.29, server: , request: "GET /api/stats HTTP/1.1", subrequest: "/auth", upstream: "http://127.0.0.1:5001/auth", host: "192.168.3.35:5000", referrer: "http://192.168.3.35:5000/"
2025-09-16 11:38:09.602788 2025/09/16 11:38:09 [error] 229#229: *1432 auth request unexpected statu

@vibr77 commented on GitHub (Sep 16, 2025): Hello, I am getting this error... 2025-09-16 11:38:09.388144 192.168.4.29 - - [16/Sep/2025:11:38:09 +0200] "GET /fonts/Inter-Medium.woff2 HTTP/1.1" 200 111380 "http://192.168.3.35:5000/" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.0 Safari/605.1.15" "-" 2025-09-16 11:38:09.602684 2025/09/16 11:38:09 [error] 229#229: *1431 connect() failed (111: Connection refused) while connecting to upstream, client: 192.168.4.29, server: , request: "GET /api/profile HTTP/1.1", subrequest: "/auth", upstream: "http://127.0.0.1:5001/auth", host: "192.168.3.35:5000", referrer: "http://192.168.3.35:5000/" 2025-09-16 11:38:09.602751 2025/09/16 11:38:09 [error] 229#229: *1431 auth request unexpected status: 502 while sending to client, client: 192.168.4.29, server: , request: "GET /api/profile HTTP/1.1", host: "192.168.3.35:5000", referrer: "http://192.168.3.35:5000/" 2025-09-16 11:38:09.602759 192.168.4.29 - - [16/Sep/2025:11:38:09 +0200] "GET /api/profile HTTP/1.1" 500 177 "http://192.168.3.35:5000/" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.0 Safari/605.1.15" "-" 2025-09-16 11:38:09.602765 2025/09/16 11:38:09 [error] 229#229: *1433 connect() failed (111: Connection refused) while connecting to upstream, client: 192.168.4.29, server: , request: "GET /api/profile HTTP/1.1", subrequest: "/auth", upstream: "http://127.0.0.1:5001/auth", host: "192.168.3.35:5000", referrer: "http://192.168.3.35:5000/" 2025-09-16 11:38:09.602771 2025/09/16 11:38:09 [error] 229#229: *1433 auth request unexpected status: 502 while sending to client, client: 192.168.4.29, server: , request: "GET /api/profile HTTP/1.1", host: "192.168.3.35:5000", referrer: "http://192.168.3.35:5000/" 2025-09-16 11:38:09.602777 192.168.4.29 - - [16/Sep/2025:11:38:09 +0200] "GET /api/profile HTTP/1.1" 500 177 "http://192.168.3.35:5000/" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.0 Safari/605.1.15" "-" 2025-09-16 11:38:09.602783 2025/09/16 11:38:09 [error] 229#229: *1432 connect() failed (111: Connection refused) while connecting to upstream, client: 192.168.4.29, server: , request: "GET /api/stats HTTP/1.1", subrequest: "/auth", upstream: "http://127.0.0.1:5001/auth", host: "192.168.3.35:5000", referrer: "http://192.168.3.35:5000/" 2025-09-16 11:38:09.602788 2025/09/16 11:38:09 [error] 229#229: *1432 auth request unexpected statu

kerem commented

Author

Owner

@vibr77 commented on GitHub (Sep 16, 2025):

and as well
root@frigate:/dev/shm/logs/frigate# tail -f current
2025-09-16 11:56:33.646438 [INFO] Preparing Frigate...
2025-09-16 11:56:33.787564 /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run: line 20: /usr/lib/ffmpeg/7.0/bin/ffmpeg: No such file or directory

@vibr77 commented on GitHub (Sep 16, 2025): and as well root@frigate:/dev/shm/logs/frigate# tail -f current 2025-09-16 11:56:33.646438 [INFO] Preparing Frigate... 2025-09-16 11:56:33.787564 /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run: line 20: /usr/lib/ffmpeg/7.0/bin/ffmpeg: No such file or directory

kerem commented

Author

Owner

@vibr77 commented on GitHub (Sep 16, 2025):

I do know if it helps, but I have nothing listening to port 5001

@vibr77 commented on GitHub (Sep 16, 2025): I do know if it helps, but I have nothing listening to port 5001

kerem commented

Author

Owner

@vibr77 commented on GitHub (Sep 16, 2025):

How do I start / stop restart manually frigate from the console ?

@vibr77 commented on GitHub (Sep 16, 2025): How do I start / stop restart manually frigate from the console ?

kerem commented

Author

Owner

@rommess commented on GitHub (Sep 16, 2025):

@vibr77 systemctl restart frigate.service or systemctl restart go2rtc.service for go2rtc

@rommess commented on GitHub (Sep 16, 2025): @vibr77 `systemctl restart frigate.service ` or `systemctl restart go2rtc.service` for go2rtc

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 16, 2025):

HW acceleration with iGPU is now functional.
I've been trying to get the Google Coral tpu working, can som" one please do a sanity-check on this config ?

root@frigate:~# cat /config/config.yml
mqtt:
  enabled: false
cameras:
  test:
    ffmpeg:
      inputs:
        - path: /media/frigate/person-bicycle-car-detection.mp4
          input_args: -re -stream_loop -1 -fflags +genpts
          roles:
            - detect
            - rtmp
    detect:
      height: 1080
      width: 1920
      fps: 5
# Optional: Authentication configuration
auth:
  # Optional: Enable authentication
  enabled: false
detect:
  enabled: true
ffmpeg:
  hwaccel_args: auto
detectors:
#  detector01:
#    type: openvino
  coral:
    type: edgetpu
    device: usb

model:
  width: 300
  height: 300
  input_tensor: nhwc
  input_pixel_format: bgr
#  path: /openvino-model/ssdlite_mobilenet_v2.xml
  path: /models/edgetpu_model.tflite
#  labelmap_path: /openvino-model/coco_91cl_bkgr.txt
  labelmap_path: /labelmap/coco.txt
version: 0.16-0
root@frigate:~# ls -l /dev/bus/usb/
total 0
drwxr-xr-x 2 root root 60 Sep 16 13:31 002
root@frigate:~#

@Zixim commented on GitHub (Sep 16, 2025): HW acceleration with iGPU is now functional. I've been trying to get the Google Coral tpu working, can som" one please do a sanity-check on this config ? ``` root@frigate:~# cat /config/config.yml mqtt: enabled: false cameras: test: ffmpeg: inputs: - path: /media/frigate/person-bicycle-car-detection.mp4 input_args: -re -stream_loop -1 -fflags +genpts roles: - detect - rtmp detect: height: 1080 width: 1920 fps: 5 # Optional: Authentication configuration auth: # Optional: Enable authentication enabled: false detect: enabled: true ffmpeg: hwaccel_args: auto detectors: # detector01: # type: openvino coral: type: edgetpu device: usb model: width: 300 height: 300 input_tensor: nhwc input_pixel_format: bgr # path: /openvino-model/ssdlite_mobilenet_v2.xml path: /models/edgetpu_model.tflite # labelmap_path: /openvino-model/coco_91cl_bkgr.txt labelmap_path: /labelmap/coco.txt version: 0.16-0 root@frigate:~# ls -l /dev/bus/usb/ total 0 drwxr-xr-x 2 root root 60 Sep 16 13:31 002 root@frigate:~# ```

kerem commented

Author

Owner

@alanoo81 commented on GitHub (Sep 16, 2025):

The edgetpu_model.tflite file seems to be missing.

2025-09-16 09:46:20.979452 ValueError: Could not open '/edgetpu_model.tflite'.

Th install script has ✔️ Downloaded Coral TPU Model which I'm assuming is the required file. Is the path for this different?

Indeed I found the model in /models/edgetpu_model.tflite
Copying it to / worked for me, although I guess you could also deal with it with model path

@alanoo81 commented on GitHub (Sep 16, 2025): > The edgetpu_model.tflite file seems to be missing. > > 2025-09-16 09:46:20.979452 ValueError: Could not open '/edgetpu_model.tflite'. > > Th install script has `✔️ Downloaded Coral TPU Model` which I'm assuming is the required file. Is the path for this different? Indeed I found the model in /models/edgetpu_model.tflite Copying it to / worked for me, although I guess you could also deal with it with model path

kerem commented

Author

Owner

@vibr77 commented on GitHub (Sep 16, 2025):

systemctl restart frigate.service

Thanks,
Do you have the same issue regarding 5001 ?
What is wrong with my install ?

@vibr77 commented on GitHub (Sep 16, 2025): > `systemctl restart frigate.service` Thanks, Do you have the same issue regarding 5001 ? What is wrong with my install ?

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 16, 2025):

@vibr77 you posted this line :
2025-09-16 11:56:33.787564 /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run: line 20: /usr/lib/ffmpeg/7.0/bin/ffmpeg

are you involving docker in this installation in some way ? This is a docker-free installation, it installs straight into an LXC.

@Zixim commented on GitHub (Sep 16, 2025): @vibr77 you posted this line : `2025-09-16 11:56:33.787564 /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run: line 20: /usr/lib/ffmpeg/7.0/bin/ffmpeg` are you involving docker in this installation in some way ? This is a docker-free installation, it installs straight into an LXC.

kerem commented

Author

Owner

@alanoo81 commented on GitHub (Sep 16, 2025):

@hawkeye217 and @NickM-27 sorry to bother you here, could you help us troubleshoot why onnx isn't leveraging iGPU? The only thing I see in the logs is: [W:onnxruntime:, graph.cc:109 MergeShapeInfo] Error merging shape info for output. 'linear_43' source:{1200,1} target:{1200}. Falling back to lenient merge. but I don't think this should be what's preventing leveraging the iGPU. Any idea as to where we should look?

On AMD though I can't make it work too.
I can see this error in Frigate logs
Message
25-09-16 14:10:43.607061453 [E:onnxruntime:Default, env.cc:234 ThreadMain] pthread_setaffinity_np failed for thread: 7606, index: 0, mask: {0, }, error code: 22 error msg: Invalid argument. Specify the number of threads explicitly so the affinity is not set.

Then it looks like it's working on CPU (so it's slow).
D-fine or yoloX gives the same error

@remz1337 for my coral tpu I had to add the passthrough for /dev/apex_0 which was missing and changing moving the model file to /

@alanoo81 commented on GitHub (Sep 16, 2025): > [@hawkeye217](https://github.com/hawkeye217) and [@NickM-27](https://github.com/NickM-27) sorry to bother you here, could you help us troubleshoot why onnx isn't leveraging iGPU? The only thing I see in the logs is: `[W:onnxruntime:, graph.cc:109 MergeShapeInfo] Error merging shape info for output. 'linear_43' source:{1200,1} target:{1200}. Falling back to lenient merge. ` but I don't think this should be what's preventing leveraging the iGPU. Any idea as to where we should look? On AMD though I can't make it work too. I can see this error in Frigate logs Message 25-09-16 14:10:43.607061453 [E:onnxruntime:Default, env.cc:234 ThreadMain] pthread_setaffinity_np failed for thread: 7606, index: 0, mask: {0, }, error code: 22 error msg: Invalid argument. Specify the number of threads explicitly so the affinity is not set. Then it looks like it's working on CPU (so it's slow). D-fine or yoloX gives the same error @remz1337 for my coral tpu I had to add the passthrough for /dev/apex_0 which was missing and changing moving the model file to /

kerem commented

Author

Owner

@timtjtim commented on GitHub (Sep 16, 2025):

https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3288364405 is happening to me, too.

  🚀  Creating a Frigate LXC using the above settings
/dev/fd/63: line 1058: FW: unbound variable
/dev/fd/63: line 1: pop_var_context: head of shell_variables not a function context

possibly related to Nvidia passthrough when I din't have an Nvidia GPU, I guess it's an unhelpful default...?

@timtjtim commented on GitHub (Sep 16, 2025): https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3288364405 is happening to me, too. ``` 🚀 Creating a Frigate LXC using the above settings /dev/fd/63: line 1058: FW: unbound variable /dev/fd/63: line 1: pop_var_context: head of shell_variables not a function context ``` possibly related to Nvidia passthrough when I din't have an Nvidia GPU, I guess it's an unhelpful default...?

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 16, 2025):

For coral users, indeed there is no automatic setup path as of now, mostly because I don't have one, so it will require some additional manual setup. I have moved the detection model for Coral at /edgetpu_model.tflite (thanks for pointing it out).

The script for now supports the following detectors:

TensorRT with Nvidia GPU (default)
Openvino with iGPU
CPU (for testing, not recommended)

I would be happy to accept a PR to officially include TPU

Also, a reminder for everyone trying this out:
bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"
for iGPU to work, you need to go into advanced, make it priviledged and do not passthrough Nvidia GPU

@remz1337 commented on GitHub (Sep 16, 2025): For coral users, indeed there is no automatic setup path as of now, mostly because I don't have one, so it will require some additional manual setup. I have moved the detection model for Coral at `/edgetpu_model.tflite` (thanks for pointing it out). The script for now supports the following detectors: - TensorRT with Nvidia GPU (default) - Openvino with iGPU - CPU (for testing, not recommended) I would be happy to accept a PR to officially include TPU Also, a reminder for everyone trying this out: `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"` for iGPU to work, you need to go into advanced, make it priviledged and do not passthrough Nvidia GPU

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 16, 2025):

@alanoo81 what did you put as labelmap_path in your frigate config.yml when using the Coral tpu?

@Zixim commented on GitHub (Sep 16, 2025): @alanoo81 what did you put as `labelmap_path` in your frigate config.yml when using the Coral tpu?

kerem commented

Author

Owner

@alanoo81 commented on GitHub (Sep 16, 2025):

@alanoo81 what did you put as labelmap_path in your frigate config.yml when using the Coral tpu?

Nothing, it was not mandatory for the edgetpu models afaik
my conf finishes like that, without any "model" section.

ffmpeg:
  hwaccel_args: preset-vaapi
detectors:
  coral:
    type: edgetpu
    device: pci
version: 0.16-0

@alanoo81 commented on GitHub (Sep 16, 2025): > [@alanoo81](https://github.com/alanoo81) what did you put as `labelmap_path` in your frigate config.yml when using the Coral tpu? Nothing, it was not mandatory for the edgetpu models afaik my conf finishes like that, without any "model" section. ``` ffmpeg: hwaccel_args: preset-vaapi detectors: coral: type: edgetpu device: pci version: 0.16-0 ```

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 16, 2025):

Also, a reminder for everyone trying this out:
bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"
for iGPU to work, you need to go into advanced, make it priviledged and do not passthrough Nvidia GPU

This worked for me on my N150 system with iGPU. Using the default config file it seems fine. It is showing a 10 ms inference and 0.32% intel-vaapi for GPU Usage.

@atrozgrifo commented on GitHub (Sep 16, 2025): > Also, a reminder for everyone trying this out: > `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"` > for iGPU to work, you need to go into advanced, make it priviledged and do not passthrough Nvidia GPU This worked for me on my N150 system with iGPU. Using the default config file it seems fine. It is showing a 10 ms inference and 0.32% intel-vaapi for GPU Usage.

kerem commented

Author

Owner

@vibr77 commented on GitHub (Sep 16, 2025):

@vibr77 you posted this line : 2025-09-16 11:56:33.787564 /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run: line 20: /usr/lib/ffmpeg/7.0/bin/ffmpeg

are you involving docker in this installation in some way ? This is a docker-free installation, it installs straight into an LXC.

my mistake, problem solved...

@vibr77 commented on GitHub (Sep 16, 2025): > [@vibr77](https://github.com/vibr77) you posted this line : `2025-09-16 11:56:33.787564 /opt/frigate/docker/main/rootfs/etc/s6-overlay/s6-rc.d/frigate/run: line 20: /usr/lib/ffmpeg/7.0/bin/ffmpeg` > > are you involving docker in this installation in some way ? This is a docker-free installation, it installs straight into an LXC. my mistake, problem solved...

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 16, 2025):

@alanoo81 sorry to get back to you, did you have to install libedgetpu1-std in the container to get the Coral up ?

@Zixim commented on GitHub (Sep 16, 2025): @alanoo81 sorry to get back to you, did you have to install `libedgetpu1-std` in the container to get the Coral up ?

kerem commented

Author

Owner

@alanoo81 commented on GitHub (Sep 16, 2025):

@alanoo81 sorry to get back to you, did you have to install libedgetpu1-std in the container to get the Coral up ?

No, did nothing more than changing the location of the model file and passing through /dev/apex_0 directly using the pve UI.

Might also depend if you are reusing a previous version conf file or not.
If not, you might have to remove the "model" section of the file too.

Although it might be worth noting that I had a working 0.14 LXC with the Coral on the same host, so drivers were already there for example.

@alanoo81 commented on GitHub (Sep 16, 2025): > [@alanoo81](https://github.com/alanoo81) sorry to get back to you, did you have to install `libedgetpu1-std` in the container to get the Coral up ? No, did nothing more than changing the location of the model file and passing through /dev/apex_0 directly using the pve UI. <img width="528" height="260" alt="Image" src="https://github.com/user-attachments/assets/dca5a4a7-79d2-46aa-b405-07fd8fd5d5c1" /> Might also depend if you are reusing a previous version conf file or not. If not, you might have to remove the "model" section of the file too. Although it might be worth noting that I had a working 0.14 LXC with the Coral on the same host, so drivers were already there for example.

kerem commented

Author

Owner

@Adrien40 commented on GitHub (Sep 16, 2025):

On Intel N150 :
bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"
for iGPU to work, you need to go into advanced, make it priviledged and do not passthrough Nvidia GPU

✔️ Downloaded OpenVino Model
⠧ Building D-FINE Model/dev/fd/63: line 223: 39620 Segmentation fault "$@" > /dev/null 2>&1
⠏ Building D-FINE Model
[ERROR] in line 224: exit code 0: while executing command "$@" > /dev/null 2>&1

[ERROR] in line 226: exit code 0: while executing command "$@" > /dev/null 2>&1

@Adrien40 commented on GitHub (Sep 16, 2025): On Intel N150 : bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)" for iGPU to work, you need to go into advanced, make it priviledged and do not passthrough Nvidia GPU ✔️ Downloaded OpenVino Model ⠧ Building D-FINE Model/dev/fd/63: line 223: 39620 Segmentation fault "$@" > /dev/null 2>&1 ⠏ Building D-FINE Model [ERROR] in line 224: exit code 0: while executing command "$@" > /dev/null 2>&1 [ERROR] in line 226: exit code 0: while executing command "$@" > /dev/null 2>&1

kerem commented

Author

Owner

@denlenin commented on GitHub (Sep 16, 2025):

mkdir -p vendor
wget -O sqlite-amalgamation.zip https://www.sqlite.org/2024/sqlite-amalgamation-3450300.zip
--2025-09-16 21:55:48-- https://www.sqlite.org/2024/sqlite-amalgamation-3450300.zip
Resolving www.sqlite.org (www.sqlite.org)... 194.195.208.62, 2600:3c02::f03c:95ff:fe07:695
Connecting to www.sqlite.org (www.sqlite.org)|194.195.208.62|:443... connected.
GnuTLS: Error in the pull function.
Unable to establish SSL connection.

[ERROR] in line 81: exit code 0: while executing command $STD bash /opt/frigate/docker/main/build_sqlite_vec.sh
/dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context

@denlenin commented on GitHub (Sep 16, 2025): + mkdir -p vendor + wget -O sqlite-amalgamation.zip https://www.sqlite.org/2024/sqlite-amalgamation-3450300.zip --2025-09-16 21:55:48-- https://www.sqlite.org/2024/sqlite-amalgamation-3450300.zip Resolving www.sqlite.org (www.sqlite.org)... 194.195.208.62, 2600:3c02::f03c:95ff:fe07:695 Connecting to www.sqlite.org (www.sqlite.org)|194.195.208.62|:443... connected. GnuTLS: Error in the pull function. Unable to establish SSL connection. [ERROR] in line 81: exit code 0: while executing command $STD bash /opt/frigate/docker/main/build_sqlite_vec.sh /dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 16, 2025):

Also, a reminder for everyone trying this out:
bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"
for iGPU to work, you need to go into advanced, make it priviledged and do not passthrough Nvidia GPU

This worked for me on my N150 system with iGPU. Using the default config file it seems fine. It is showing a 10 ms inference and 0.32% intel-vaapi for GPU Usage.

It may not be working well. I added 2 cameras (that were previously used on a Coral.AI supported system). 1 camera causes frigate to crash and restart continuously, the other seems to run fine for 3 minutes or so then I get errors like this:

2025-09-16 11:56:29.643962  [2025-09-16 11:56:29] frigate.video                  ERROR   : Driveway: Unable to read frames from ffmpeg process.
2025-09-16 11:56:29.644741  [2025-09-16 11:56:29] frigate.video                  ERROR   : Driveway: ffmpeg process is not running. exiting capture thread...
2025-09-16 11:56:34.784166  [2025-09-16 11:56:34] watchdog.Driveway              ERROR   : Ffmpeg process crashed unexpectedly for Driveway.
2025-09-16 11:56:34.784553  [2025-09-16 11:56:34] watchdog.Driveway              ERROR   : The following ffmpeg logs include the last 100 lines prior to exit.
2025-09-16 11:56:34.784954  [2025-09-16 11:56:34] ffmpeg.Driveway.detect         ERROR   : [AVHWFramesContext @ 0x708f480aefc0] Failed to sync surface 0x13: 1 (operation failed).
2025-09-16 11:56:34.785317  [2025-09-16 11:56:34] ffmpeg.Driveway.detect         ERROR   : [hwdownload @ 0x708f50003980] Failed to download frame: -5.
2025-09-16 11:56:34.785639  [2025-09-16 11:56:34] ffmpeg.Driveway.detect         ERROR   : [vf#0:0 @ 0x5892c2b9b300] Error while filtering: Input/output error
2025-09-16 11:56:34.785910  [2025-09-16 11:56:34] ffmpeg.Driveway.detect         ERROR   : [vf#0:0 @ 0x5892c2b9b300] Task finished with error code: -5 (Input/output error)
2025-09-16 11:56:34.786173  [2025-09-16 11:56:34] ffmpeg.Driveway.detect         ERROR   : [vf#0:0 @ 0x5892c2b9b300] Terminating thread with return code -5 (Input/output error)

@atrozgrifo commented on GitHub (Sep 16, 2025): > > Also, a reminder for everyone trying this out: > > `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"` > > for iGPU to work, you need to go into advanced, make it priviledged and do not passthrough Nvidia GPU > > This worked for me on my N150 system with iGPU. Using the default config file it seems fine. It is showing a 10 ms inference and 0.32% intel-vaapi for GPU Usage. It may not be working well. I added 2 cameras (that were previously used on a Coral.AI supported system). 1 camera causes frigate to crash and restart continuously, the other seems to run fine for 3 minutes or so then I get errors like this: ``` 2025-09-16 11:56:29.643962 [2025-09-16 11:56:29] frigate.video ERROR : Driveway: Unable to read frames from ffmpeg process. 2025-09-16 11:56:29.644741 [2025-09-16 11:56:29] frigate.video ERROR : Driveway: ffmpeg process is not running. exiting capture thread... 2025-09-16 11:56:34.784166 [2025-09-16 11:56:34] watchdog.Driveway ERROR : Ffmpeg process crashed unexpectedly for Driveway. 2025-09-16 11:56:34.784553 [2025-09-16 11:56:34] watchdog.Driveway ERROR : The following ffmpeg logs include the last 100 lines prior to exit. 2025-09-16 11:56:34.784954 [2025-09-16 11:56:34] ffmpeg.Driveway.detect ERROR : [AVHWFramesContext @ 0x708f480aefc0] Failed to sync surface 0x13: 1 (operation failed). 2025-09-16 11:56:34.785317 [2025-09-16 11:56:34] ffmpeg.Driveway.detect ERROR : [hwdownload @ 0x708f50003980] Failed to download frame: -5. 2025-09-16 11:56:34.785639 [2025-09-16 11:56:34] ffmpeg.Driveway.detect ERROR : [vf#0:0 @ 0x5892c2b9b300] Error while filtering: Input/output error 2025-09-16 11:56:34.785910 [2025-09-16 11:56:34] ffmpeg.Driveway.detect ERROR : [vf#0:0 @ 0x5892c2b9b300] Task finished with error code: -5 (Input/output error) 2025-09-16 11:56:34.786173 [2025-09-16 11:56:34] ffmpeg.Driveway.detect ERROR : [vf#0:0 @ 0x5892c2b9b300] Terminating thread with return code -5 (Input/output error) ```

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 16, 2025):

@atrozgrifo can you enable detection with the default test video in loop? (detect: true) Then paste the full frigate logs. I suspect something is wrong with ffmpeg leveraging hwaccel

@remz1337 commented on GitHub (Sep 16, 2025): @atrozgrifo can you enable detection with the default test video in loop? (detect: true) Then paste the full frigate logs. I suspect something is wrong with ffmpeg leveraging hwaccel

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 16, 2025):

On Intel N150 : bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)" for iGPU to work, you need to go into advanced, make it priviledged and do not passthrough Nvidia GPU

✔️ Downloaded OpenVino Model ⠧ Building D-FINE Model/dev/fd/63: line 223: 39620 Segmentation fault "$@" > /dev/null 2>&1 ⠏ Building D-FINE Model [ERROR] in line 224: exit code 0: while executing command "$@" > /dev/null 2>&1

[ERROR] in line 226: exit code 0: while executing command "$@" > /dev/null 2>&1

don't worry about the D-FINE error, it's an alternate model I was testing. It still compiles correctly

@remz1337 commented on GitHub (Sep 16, 2025): > On Intel N150 : bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)" for iGPU to work, you need to go into advanced, make it priviledged and do not passthrough Nvidia GPU > > ✔️ Downloaded OpenVino Model ⠧ Building D-FINE Model/dev/fd/63: line 223: 39620 Segmentation fault "$@" > /dev/null 2>&1 ⠏ Building D-FINE Model [ERROR] in line 224: exit code 0: while executing command "$@" > /dev/null 2>&1 > > [ERROR] in line 226: exit code 0: while executing command "$@" > /dev/null 2>&1 don't worry about the D-FINE error, it's an alternate model I was testing. It still compiles correctly

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 16, 2025):

@atrozgrifo can you enable detection with the default test video in loop? (detect: true) Then paste the full frigate logs. I suspect something is wrong with ffmpeg leveraging hwaccel

I didn't save the original configuration so I had to recreate it. This is what I have:

mqtt:
  host: mqtt

detect:
  enabled: true
ffmpeg:
  hwaccel_args: auto

detectors:
  detector01:
    type: openvino
model:
  width: 300
  height: 300
  input_tensor: nhwc
  input_pixel_format: bgr
  path: /openvino-model/ssdlite_mobilenet_v2.xml
  labelmap_path: /openvino-model/coco_91cl_bkgr.txt

cameras:
  test:
    ffmpeg:
      inputs:
        - path: /media/frigate/person-bicycle-car-detection.mp4
          input_args: -re -stream_loop -1 -fflags +genpts
          roles:
            - detect
            - rtmp
    detect:
      height: 1080
      width: 1920
      fps: 5
version: 0.16-0

Log is showing me this:

root@frigate16node1:/config# tail -f  /dev/shm/logs/frigate/current 
2025-09-16 16:33:09.250059  [2025-09-16 16:33:09] peewee_migrate.logs            INFO    : There is nothing to migrate
2025-09-16 16:33:09.275682  [2025-09-16 16:33:09] frigate.app                    INFO    : Recording process started: 838
2025-09-16 16:33:09.276338  [2025-09-16 16:33:09] frigate.app                    INFO    : Review process started: 851
2025-09-16 16:33:09.278519  [2025-09-16 16:33:09] frigate.app                    INFO    : go2rtc process pid: 158
2025-09-16 16:33:09.296627  [2025-09-16 16:33:09] detector.detector01            INFO    : Starting detection process: 863
2025-09-16 16:33:09.314332  [2025-09-16 16:33:09] frigate.app                    INFO    : Output process started: 880
2025-09-16 16:33:09.358128  [2025-09-16 16:33:09] frigate.app                    INFO    : Camera processor started for test: 903
2025-09-16 16:33:09.386477  [2025-09-16 16:33:09] frigate.app                    INFO    : Capture process started for test: 907
2025-09-16 16:33:09.731418  [2025-09-16 16:33:09] frigate.api.fastapi_app        INFO    : Starting FastAPI app
2025-09-16 16:33:09.849339  [2025-09-16 16:33:09] frigate.api.fastapi_app        INFO    : FastAPI started

But my Detectors stats are:

Up to 72% CPU usage.

@atrozgrifo commented on GitHub (Sep 16, 2025): > [@atrozgrifo](https://github.com/atrozgrifo) can you enable detection with the default test video in loop? (detect: true) Then paste the full frigate logs. I suspect something is wrong with ffmpeg leveraging hwaccel I didn't save the original configuration so I had to recreate it. This is what I have: ``` mqtt: host: mqtt detect: enabled: true ffmpeg: hwaccel_args: auto detectors: detector01: type: openvino model: width: 300 height: 300 input_tensor: nhwc input_pixel_format: bgr path: /openvino-model/ssdlite_mobilenet_v2.xml labelmap_path: /openvino-model/coco_91cl_bkgr.txt cameras: test: ffmpeg: inputs: - path: /media/frigate/person-bicycle-car-detection.mp4 input_args: -re -stream_loop -1 -fflags +genpts roles: - detect - rtmp detect: height: 1080 width: 1920 fps: 5 version: 0.16-0 ``` Log is showing me this: ``` root@frigate16node1:/config# tail -f /dev/shm/logs/frigate/current 2025-09-16 16:33:09.250059 [2025-09-16 16:33:09] peewee_migrate.logs INFO : There is nothing to migrate 2025-09-16 16:33:09.275682 [2025-09-16 16:33:09] frigate.app INFO : Recording process started: 838 2025-09-16 16:33:09.276338 [2025-09-16 16:33:09] frigate.app INFO : Review process started: 851 2025-09-16 16:33:09.278519 [2025-09-16 16:33:09] frigate.app INFO : go2rtc process pid: 158 2025-09-16 16:33:09.296627 [2025-09-16 16:33:09] detector.detector01 INFO : Starting detection process: 863 2025-09-16 16:33:09.314332 [2025-09-16 16:33:09] frigate.app INFO : Output process started: 880 2025-09-16 16:33:09.358128 [2025-09-16 16:33:09] frigate.app INFO : Camera processor started for test: 903 2025-09-16 16:33:09.386477 [2025-09-16 16:33:09] frigate.app INFO : Capture process started for test: 907 2025-09-16 16:33:09.731418 [2025-09-16 16:33:09] frigate.api.fastapi_app INFO : Starting FastAPI app 2025-09-16 16:33:09.849339 [2025-09-16 16:33:09] frigate.api.fastapi_app INFO : FastAPI started ``` But my Detectors stats are: <img width="1459" height="555" alt="Image" src="https://github.com/user-attachments/assets/908f2dfa-227b-4b49-af81-671948663a06" /> Up to 72% CPU usage.

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 16, 2025):

I just realized that I didn't have detect enabled on my original one with the camera and ffmpeg crashing.

I decided to enable the detect in my camera config one and add in the video file 'camera'.

It's back to periodic crashing. I have 2 detectors going because I wasn't sure if having just one was causing the crashing with 2 feeds.

@atrozgrifo commented on GitHub (Sep 16, 2025): I just realized that I didn't have detect enabled on my original one with the camera and ffmpeg crashing. I decided to enable the detect in my camera config one and add in the video file 'camera'. It's back to periodic crashing. I have 2 detectors going because I wasn't sure if having just one was causing the crashing with 2 feeds. <img width="1459" height="750" alt="Image" src="https://github.com/user-attachments/assets/ad21b572-fc33-4030-92a1-df5ffbc910b2" />

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 17, 2025):

iGPU seems to be working fine. I've heard from the frigate devs that v16 uses more detector CPU than anticipated and that it should be improved in v17. Regarding the random crashes, try increasing the RAM, SWAP and CPU count. Let's see if it some kind of memory leak.

Additionnaly, look at the frigate doc. Might be worth trying out other hwaccel presets than auto:
https://docs.frigate.video/configuration/hardware_acceleration_video#intel-based-cpus

@remz1337 commented on GitHub (Sep 17, 2025): iGPU seems to be working fine. I've heard from the frigate devs that v16 uses more detector CPU than anticipated and that it should be improved in v17. Regarding the random crashes, try increasing the RAM, SWAP and CPU count. Let's see if it some kind of memory leak. Additionnaly, look at the frigate doc. Might be worth trying out other hwaccel presets than auto: https://docs.frigate.video/configuration/hardware_acceleration_video#intel-based-cpus

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 17, 2025):

iGPU seems to be working fine. I've heard from the frigate devs that v16 uses more detector CPU than anticipated and that it should be improved in v17. Regarding the random crashes, try increasing the RAM, SWAP and CPU count. Let's see if it some kind of memory leak.

Additionnaly, look at the frigate doc. Might be worth trying out other hwaccel presets than auto: https://docs.frigate.video/configuration/hardware_acceleration_video#intel-based-cpus

I went from the default RAM, SWAP and CPU to max on my little N150 (4 core, 4G swap and 8G RAM). Ran for nearly 10 minutes with no crash. High CPU usage still.

Started messing around with the hwaccel presets. I tried "hwaccel_args: preset-intel-qsv-h264" and that works for my camera but it's not the right decoder for the sample video file. I tried *h265 as well and that also didn't work. With just the 1 camera my CPU usage is way down in Proxmox. In frigate is is also down but I don't see any GPU usage, not sure if the iGPU should show there or not, I was getting some values there before. It even says "intel-qsv" on it.

I added a 2nd camera with the *h264 and it seems to be working. This is a camera I mentioned earlier that would cause an instant crash.

if I set the test file to hwaccel_args: auto a lot of CPU time gets used on it. If I leave it as a *h264 which doesn't work, my CPU usage drops quite a lot. I've now commented out the test video as I don't know what decoder to use but Auto which seems to be using the CPU.

I dropped my hardware to 3 processors and 4 gigs of RAM so that I wasn't using everything. My CPU usage with the 2 cameras is sitting about 15% +-3% according to Proxmox. RAM usage is about 1.76 GB.

Log files seem pretty clean.

I added in 2 more cameras, started having problems again, ffmpeg seems to be crashing again. Went back to 3 cameras. CPU usage has climbed a few %. RAM usage is now at 1.96 (up 0.20) GB.

intel_gpu_top is showing a lot more activity than the other day.

I need sleep, stopping for now.

@atrozgrifo commented on GitHub (Sep 17, 2025): > iGPU seems to be working fine. I've heard from the frigate devs that v16 uses more detector CPU than anticipated and that it should be improved in v17. Regarding the random crashes, try increasing the RAM, SWAP and CPU count. Let's see if it some kind of memory leak. > > Additionnaly, look at the frigate doc. Might be worth trying out other hwaccel presets than auto: https://docs.frigate.video/configuration/hardware_acceleration_video#intel-based-cpus I went from the default RAM, SWAP and CPU to max on my little N150 (4 core, 4G swap and 8G RAM). Ran for nearly 10 minutes with no crash. High CPU usage still. Started messing around with the hwaccel presets. I tried "hwaccel_args: preset-intel-qsv-h264" and that works for my camera but it's not the right decoder for the sample video file. I tried *h265 as well and that also didn't work. With just the 1 camera my CPU usage is way down in Proxmox. In frigate is is also down but I don't see any GPU usage, not sure if the iGPU should show there or not, I was getting some values there before. It even says "intel-qsv" on it. I added a 2nd camera with the *h264 and it seems to be working. This is a camera I mentioned earlier that would cause an instant crash. if I set the test file to hwaccel_args: auto a lot of CPU time gets used on it. If I leave it as a *h264 which doesn't work, my CPU usage drops quite a lot. I've now commented out the test video as I don't know what decoder to use but Auto which seems to be using the CPU. I dropped my hardware to 3 processors and 4 gigs of RAM so that I wasn't using everything. My CPU usage with the 2 cameras is sitting about 15% +-3% according to Proxmox. RAM usage is about 1.76 GB. Log files seem pretty clean. I added in 2 more cameras, started having problems again, ffmpeg seems to be crashing again. Went back to 3 cameras. CPU usage has climbed a few %. RAM usage is now at 1.96 (up 0.20) GB. intel_gpu_top is showing a lot more activity than the other day. <img width="1492" height="278" alt="Image" src="https://github.com/user-attachments/assets/9f8f915c-ebdc-492d-9a69-5a5c9b129233" /> I need sleep, stopping for now.

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 17, 2025):

Re-installed the latest version,

Added 1 line to thc lxc conf :

# Coral
lxc.mount.entry: /dev/bus/usb/002/ dev/bus/usb/002/ none bind,optional,create=dir 0,0

Frigate config :

detect:
  enabled: true
ffmpeg:
  hwaccel_args: auto
detectors:
  coral:
    type: edgetpu
    device: usb
version: 0.16-0

The logging for frigate shows TPU found, good news :

2025-09-17 12:51:17.958649  [2025-09-17 12:51:17] frigate.detectors.plugins.edgetpu_tfl INFO    : TPU found
2025-09-17 12:51:17.962067  INFO: Created TensorFlow Lite XNNPACK delegate for CPU.

but the web interface shows high CPU usage, and Intel GPU :

Ideas anyone please ?

@Zixim commented on GitHub (Sep 17, 2025): Re-installed the latest version, Added 1 line to thc lxc conf : ``` # Coral lxc.mount.entry: /dev/bus/usb/002/ dev/bus/usb/002/ none bind,optional,create=dir 0,0 ``` Frigate config : ``` detect: enabled: true ffmpeg: hwaccel_args: auto detectors: coral: type: edgetpu device: usb version: 0.16-0 ``` The logging for frigate shows TPU found, good news : ``` 2025-09-17 12:51:17.958649 [2025-09-17 12:51:17] frigate.detectors.plugins.edgetpu_tfl INFO : TPU found 2025-09-17 12:51:17.962067 INFO: Created TensorFlow Lite XNNPACK delegate for CPU. ``` but the web interface shows high CPU usage, and `Intel GPU` : <img width="229" height="103" alt="Image" src="https://github.com/user-attachments/assets/8d528dac-4b59-4fef-87ab-e72dd4fa0aa6" /> Ideas anyone please ?

kerem commented

Author

Owner

@remz1337 commented on GitHub (Sep 17, 2025):

Can you try with the official Docker image? I'm curious to see if you would get the same issues with Docker.

@remz1337 commented on GitHub (Sep 17, 2025): Can you try with the official Docker image? I'm curious to see if you would get the same issues with Docker.

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 17, 2025):

I cloned the LXC that I had running on the N150 and moved it to the i5 system, left it with OpenVino support for detect. Keeps crashing. Replaced OpenVino with the Coral.AI PCI and it seems stable. I commented out the model stuff because it seemed to be the wrong model for Coral.

@atrozgrifo commented on GitHub (Sep 17, 2025): I cloned the LXC that I had running on the N150 and moved it to the i5 system, left it with OpenVino support for detect. Keeps crashing. Replaced OpenVino with the Coral.AI PCI and it seems stable. I commented out the model stuff because it seemed to be the wrong model for Coral.

kerem commented

Author

Owner

@fsrforums commented on GitHub (Sep 17, 2025):

On Intel N150 : bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)" for iGPU to work, you need to go into advanced, make it priviledged and do not passthrough Nvidia GPU
✔️ Downloaded OpenVino Model ⠧ Building D-FINE Model/dev/fd/63: line 223: 39620 Segmentation fault "$@" > /dev/null 2>&1 ⠏ Building D-FINE Model [ERROR] in line 224: exit code 0: while executing command "$@" > /dev/null 2>&1
[ERROR] in line 226: exit code 0: while executing command "$@" > /dev/null 2>&1

don't worry about the D-FINE error, it's an alternate model I was testing. It still compiles correctly

Hi team!! First comment on Github ,-( ... On MS-01 Intel 13900H

✔️ Built HailoRT
✔️ Installed OpenVino Runtime and Dev library
⠧ Downloading OpenVino Model
[ERROR] in line 195: exit code 0: while executing command wget -q https://github.com/openvinotoolkit/open_model_zoo/raw/master/data/dataset_classes/coco_91cl_bkgr.txt -O /openvino-model/coco_91cl_bkgr.txt
/dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context

@fsrforums commented on GitHub (Sep 17, 2025): > > On Intel N150 : bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)" for iGPU to work, you need to go into advanced, make it priviledged and do not passthrough Nvidia GPU > > ✔️ Downloaded OpenVino Model ⠧ Building D-FINE Model/dev/fd/63: line 223: 39620 Segmentation fault "$@" > /dev/null 2>&1 ⠏ Building D-FINE Model [ERROR] in line 224: exit code 0: while executing command "$@" > /dev/null 2>&1 > > [ERROR] in line 226: exit code 0: while executing command "$@" > /dev/null 2>&1 > > don't worry about the D-FINE error, it's an alternate model I was testing. It still compiles correctly Hi team!! First comment on Github ,-( ... On MS-01 Intel 13900H ✔️ Built HailoRT ✔️ Installed OpenVino Runtime and Dev library ⠧ Downloading OpenVino Model [ERROR] in line 195: exit code 0: while executing command wget -q https://github.com/openvinotoolkit/open_model_zoo/raw/master/data/dataset_classes/coco_91cl_bkgr.txt -O /openvino-model/coco_91cl_bkgr.txt /dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context

kerem commented

Author

Owner

@tom0010 commented on GitHub (Sep 18, 2025):

For coral users, indeed there is no automatic setup path as of now, mostly because I don't have one, so it will require some additional manual setup. I have moved the detection model for Coral at /edgetpu_model.tflite (thanks for pointing it out).

The script for now supports the following detectors:

TensorRT with Nvidia GPU (default)

Openvino with iGPU

CPU (for testing, not recommended)

I would be happy to accept a PR to officially include TPU

Also, a reminder for everyone trying this out: bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)" for iGPU to work, you need to go into advanced, make it priviledged and do not passthrough Nvidia GPU

Got this working using a Coral, the installation did take a while compared to previous installations.
But after I booted it up, I added my config from frigate 14 and changed my lxc config to enable the USB Coral:

lxc.cgroup2.devices.allow: c 189:* rwm # usb coral
lxc.mount.entry: /dev/bus/usb/002 dev/bus/usb/002 none bind,optional,create=dir #extra

My container id is 203 so these lines were added below in /etc/pve/lxc/203.conf.

I have noticed that go2rtc is not taking in my camera, so I had to specifically skip that for now and add the direct camera stream inside of the camera config section.

CPU idling around 10% right now, Coral on 7/8ms interference as before.

Using N100 CPU.

@tom0010 commented on GitHub (Sep 18, 2025): > For coral users, indeed there is no automatic setup path as of now, mostly because I don't have one, so it will require some additional manual setup. I have moved the detection model for Coral at `/edgetpu_model.tflite` (thanks for pointing it out). > > The script for now supports the following detectors: > > * TensorRT with Nvidia GPU (default) > * Openvino with iGPU > * CPU (for testing, not recommended) > > I would be happy to accept a PR to officially include TPU > > Also, a reminder for everyone trying this out: `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"` for iGPU to work, you need to go into advanced, make it priviledged and do not passthrough Nvidia GPU Got this working using a Coral, the installation did take a while compared to previous installations. But after I booted it up, I added my config from frigate 14 and changed my lxc config to enable the USB Coral: ```bash lxc.cgroup2.devices.allow: c 189:* rwm # usb coral lxc.mount.entry: /dev/bus/usb/002 dev/bus/usb/002 none bind,optional,create=dir #extra ``` My container id is 203 so these lines were added below in `/etc/pve/lxc/203.conf`. I have noticed that go2rtc is not taking in my camera, so I had to specifically skip that for now and add the direct camera stream inside of the camera config section. CPU idling around 10% right now, Coral on 7/8ms interference as before. Using N100 CPU.

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 18, 2025):

I must have been chasing ghosts for a few days :

CPU and intel GPU but nothing about TPU/Coral on the main page:

And in the System Mertics page there is Coral :

Any one else, with a Coral, seeing the same ?

@Zixim commented on GitHub (Sep 18, 2025): I must have been chasing ghosts for a few days : CPU and intel GPU but nothing about TPU/Coral on the main page: <img width="298" height="128" alt="Image" src="https://github.com/user-attachments/assets/0ae41ea6-69e7-4af5-8790-eca96bdd7131" /> And in the System Mertics page there is Coral : <img width="771" height="358" alt="Image" src="https://github.com/user-attachments/assets/d5802268-7d9a-4300-81be-1d1739453cd2" /> Any one else, with a Coral, seeing the same ?

kerem commented

Author

Owner

@Aareon commented on GitHub (Sep 18, 2025):

    ______     _             __     
   / ____/____(_)___ _____ _/ /____ 
  / /_  / ___/ / __ `/ __ `/ __/ _ \
 / __/ / /  / / /_/ / /_/ / /_/  __/
/_/   /_/  /_/\__, /\__,_/\__/\___/ 
             /____/                 
  ⚙️  Using Default Settings on node prodesk1
  🆔  Container ID: 135
  🖥️  Operating System: debian (11)
  📦  Container Type: Privileged
  💾  Disk Size: 20 GB
  🧠  CPU Cores: 4
  🛠️  RAM Size: 4096 MiB
  🚀  Creating a Frigate LXC using the above default settings
  
  ✔️   Storage local (Free: 75.9GB  Used: 11.2GB) [Template]
  ✔️   Storage local-lvm (Free: 143.8GB  Used: 197.0GB) [Container]
  ✔️   Cluster is quorate
  ✔️   Template debian-11-standard_11.7-1_amd64.tar.zst [online]
   💡   Template debian-11-standard_11.7-1_amd64.tar.zst is missing or corrupted. Re-downloading.
  ✔️   Template download successful.
  ✔️   LXC Container 135 was successfully created.

 ⚙️  Configuring VAAPI passthrough for LXC container
 ℹ️  VAAPI enables GPU hardware acceleration (e.g., for video transcoding in Jellyfin or Plex).

➤ Automatically mount all available VAAPI devices? [Y/n]: 
  ✔️   Started LXC Container
  ✔️   Network in LXC is reachable (ping)
  ✔️   Customized LXC Container
  ✔️   Set up Container OS
  ✔️   Network Connected: 10.0.0.199 
  ✔️   IPv4 Internet Connected
   ✖️   IPv6 Internet Not Connected
  ✔️   Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) git.community-scripts.org:(✔️ )
  ✔️   Updated Container OS
  ✔️   Installed Dependencies
  ✔️   Setup Python3
  ✔️   Setup Node.js 22
  ✔️   Installed go2rtc
  ✔️   Set Up Hardware Acceleration
⠙ Installing Frigate v0.14.1 (Perseverance)sleep: /lib64/ld-linux-x86-64.so.2: version `GLIBC_2.35' not found (required by /lib/x86_64-linux-gnu/libc.so.6)
curl: /lib64/ld-linux-x86-64.so.2: version `GLIBC_2.35' not found (required by /lib/x86_64-linux-gnu/libc.so.6)

[ERROR] in line 278: exit code 0: while executing command sleep 0.1
/dev/stdin: line 43: post_update_to_api: command not found

@Aareon commented on GitHub (Sep 18, 2025): ``` ______ _ __ / ____/____(_)___ _____ _/ /____ / /_ / ___/ / __ `/ __ `/ __/ _ \ / __/ / / / / /_/ / /_/ / /_/ __/ /_/ /_/ /_/\__, /\__,_/\__/\___/ /____/ ⚙️ Using Default Settings on node prodesk1 🆔 Container ID: 135 🖥️ Operating System: debian (11) 📦 Container Type: Privileged 💾 Disk Size: 20 GB 🧠 CPU Cores: 4 🛠️ RAM Size: 4096 MiB 🚀 Creating a Frigate LXC using the above default settings ✔️ Storage local (Free: 75.9GB Used: 11.2GB) [Template] ✔️ Storage local-lvm (Free: 143.8GB Used: 197.0GB) [Container] ✔️ Cluster is quorate ✔️ Template debian-11-standard_11.7-1_amd64.tar.zst [online] 💡 Template debian-11-standard_11.7-1_amd64.tar.zst is missing or corrupted. Re-downloading. ✔️ Template download successful. ✔️ LXC Container 135 was successfully created. ⚙️ Configuring VAAPI passthrough for LXC container ℹ️ VAAPI enables GPU hardware acceleration (e.g., for video transcoding in Jellyfin or Plex). ➤ Automatically mount all available VAAPI devices? [Y/n]: ✔️ Started LXC Container ✔️ Network in LXC is reachable (ping) ✔️ Customized LXC Container ✔️ Set up Container OS ✔️ Network Connected: 10.0.0.199 ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) git.community-scripts.org:(✔️ ) ✔️ Updated Container OS ✔️ Installed Dependencies ✔️ Setup Python3 ✔️ Setup Node.js 22 ✔️ Installed go2rtc ✔️ Set Up Hardware Acceleration ⠙ Installing Frigate v0.14.1 (Perseverance)sleep: /lib64/ld-linux-x86-64.so.2: version `GLIBC_2.35' not found (required by /lib/x86_64-linux-gnu/libc.so.6) curl: /lib64/ld-linux-x86-64.so.2: version `GLIBC_2.35' not found (required by /lib/x86_64-linux-gnu/libc.so.6) [ERROR] in line 278: exit code 0: while executing command sleep 0.1 /dev/stdin: line 43: post_update_to_api: command not found ```

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 18, 2025):

Installing Frigate v0.14.1

Are you using the right script? Should be 0.16.1 not 14

@atrozgrifo commented on GitHub (Sep 18, 2025): > Installing Frigate v0.14.1 Are you using the right script? Should be 0.16.1 not 14

kerem commented

Author

Owner

@Aareon commented on GitHub (Sep 18, 2025):

Whichever script was linked on the website, attempted install today via
install command

On Thu, Sep 18, 2025, 12:41 PM atrozgrifo @.***> wrote:

atrozgrifo left a comment (community-scripts/ProxmoxVE#6915)
https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3308754823

Installing Frigate v0.14.1

Are you using the right script? Should be 0.16.1 not 14

—
Reply to this email directly, view it on GitHub
https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3308754823,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/ACFLDZXEGIR3RHAKDXL2EML3TLVEBAVCNFSM6AAAAACEC6ZHIWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTGMBYG42TIOBSGM
.
You are receiving this because you commented.Message ID:
@.***>

@Aareon commented on GitHub (Sep 18, 2025): Whichever script was linked on the website, attempted install today via install command On Thu, Sep 18, 2025, 12:41 PM atrozgrifo ***@***.***> wrote: > *atrozgrifo* left a comment (community-scripts/ProxmoxVE#6915) > <https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3308754823> > > Installing Frigate v0.14.1 > > Are you using the right script? Should be 0.16.1 not 14 > > — > Reply to this email directly, view it on GitHub > <https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3308754823>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ACFLDZXEGIR3RHAKDXL2EML3TLVEBAVCNFSM6AAAAACEC6ZHIWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTGMBYG42TIOBSGM> > . > You are receiving this because you commented.Message ID: > ***@***.***> >

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Sep 18, 2025):

This issue is tracking the development of a new script for Frigate 0.16.x. There may be better issue already open for that script or you may need to open one.

@atrozgrifo commented on GitHub (Sep 18, 2025): This issue is tracking the development of a new script for Frigate 0.16.x. There may be better issue already open for that script or you may need to open one.

kerem commented

Author

Owner

@di5cord20 commented on GitHub (Sep 18, 2025):

For coral users, indeed there is no automatic setup path as of now, mostly because I don't have one, so it will require some additional manual setup. I have moved the detection model for Coral at /edgetpu_model.tflite (thanks for pointing it out).
The script for now supports the following detectors:

TensorRT with Nvidia GPU (default)

Openvino with iGPU

CPU (for testing, not recommended)

I would be happy to accept a PR to officially include TPU
Also, a reminder for everyone trying this out: bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)" for iGPU to work, you need to go into advanced, make it priviledged and do not passthrough Nvidia GPU

Got this working using a Coral, the installation did take a while compared to previous installations. But after I booted it up, I added my config from frigate 14 and changed my lxc config to enable the USB Coral:

lxc.cgroup2.devices.allow: c 189:* rwm # usb coral
lxc.mount.entry: /dev/bus/usb/002 dev/bus/usb/002 none bind,optional,create=dir #extra
My container id is 203 so these lines were added below in /etc/pve/lxc/203.conf.

I have noticed that go2rtc is not taking in my camera, so I had to specifically skip that for now and add the direct camera stream inside of the camera config section.

CPU idling around 10% right now, Coral on 7/8ms interference as before.

Using N100 CPU.

I'm about to give this a try. Am also running on N100 with a Coral USB. To clarify, are these the correct steps?

Execute install script in proxmox host shell
bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"

Question - do I need to go into advanced, make it priviledged and do not passthrough Nvidia GPU?

Update the conf file for the created LXC with:

lxc.cgroup2.devices.allow: c 189:* rwm # usb coral
lxc.mount.entry: /dev/bus/usb/002 dev/bus/usb/002 none bind,optional,create=dir #extra

Update frigate config

@di5cord20 commented on GitHub (Sep 18, 2025): > > For coral users, indeed there is no automatic setup path as of now, mostly because I don't have one, so it will require some additional manual setup. I have moved the detection model for Coral at `/edgetpu_model.tflite` (thanks for pointing it out). > > The script for now supports the following detectors: > > > > * TensorRT with Nvidia GPU (default) > > * Openvino with iGPU > > * CPU (for testing, not recommended) > > > > I would be happy to accept a PR to officially include TPU > > Also, a reminder for everyone trying this out: `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"` for iGPU to work, you need to go into advanced, make it priviledged and do not passthrough Nvidia GPU > > Got this working using a Coral, the installation did take a while compared to previous installations. But after I booted it up, I added my config from frigate 14 and changed my lxc config to enable the USB Coral: > > lxc.cgroup2.devices.allow: c 189:* rwm # usb coral > lxc.mount.entry: /dev/bus/usb/002 dev/bus/usb/002 none bind,optional,create=dir #extra > My container id is 203 so these lines were added below in `/etc/pve/lxc/203.conf`. > > I have noticed that go2rtc is not taking in my camera, so I had to specifically skip that for now and add the direct camera stream inside of the camera config section. > > CPU idling around 10% right now, Coral on 7/8ms interference as before. > > Using N100 CPU. I'm about to give this a try. Am also running on N100 with a Coral USB. To clarify, are these the correct steps? 1. Execute install script in proxmox host shell ```bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"``` **Question** - do I need to go into advanced, make it priviledged and do not passthrough Nvidia GPU? 2. Update the conf file for the created LXC with: ``` lxc.cgroup2.devices.allow: c 189:* rwm # usb coral lxc.mount.entry: /dev/bus/usb/002 dev/bus/usb/002 none bind,optional,create=dir #extra ``` 3. Update frigate config

kerem commented

Author

Owner

@crhy commented on GitHub (Sep 18, 2025):

I have a first draft working (barebone install, no hwaccel, just the sample video playing in loop). Would appreciate if some could test and report any issue. I'll add hwaccel soon

bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/main/ct/frigate.sh)"

Didn't work for me. The current frigate.sh also doesn't work for me. I opened an issue. Here's the error:

⠧ No local template found, checking online repository0: line 287: TEMPLATES: bad array subscript
⠙ No local template found, checking online repository
[ERROR] in line 1152: exit code 0: while executing command bash -c "(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/misc/create_lxc.sh)" ?

⠹ No local template found, checking online repository

@crhy commented on GitHub (Sep 18, 2025): > I have a first draft working (barebone install, no hwaccel, just the sample video playing in loop). Would appreciate if some could test and report any issue. I'll add hwaccel soon > > `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/main/ct/frigate.sh)"` Didn't work for me. The current frigate.sh also doesn't work for me. I opened an issue. Here's the error: ⠧ No local template found, checking online repository0: line 287: TEMPLATES: bad array subscript ⠙ No local template found, checking online repository [ERROR] in line 1152: exit code 0: while executing command bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/misc/create_lxc.sh)" $? ⠹ No local template found, checking online repository

kerem commented

Author

Owner

@SubNoizey commented on GitHub (Sep 19, 2025):

3. Update frigate config

how'd you go? I got the lxc up and running following your steps but I'm having issues with loading my tpu, it won't find it. keeps throwing errors

from host - "lsusb"
Bus 003 Device 003: ID 18d1:9302 Google Inc.

from frigate lxc- "ls -la /dev/bus/usb/003/"

total 0
drwxr-xr-x 2 root root          80 Sep 19 13:56 .
drwxr-xr-x 3 root root          60 Sep 19 15:52 ..
crw-rw-r-- 1 root root    189, 256 Sep 18 13:19 001
crw-rw---- 1 root plugdev 189, 258 Sep 19 15:54 003

@SubNoizey commented on GitHub (Sep 19, 2025): > 3\. Update frigate config how'd you go? I got the lxc up and running following your steps but I'm having issues with loading my tpu, it won't find it. keeps throwing errors from host - "lsusb" `Bus 003 Device 003: ID 18d1:9302 Google Inc.` from frigate lxc- "ls -la /dev/bus/usb/003/" ``` total 0 drwxr-xr-x 2 root root 80 Sep 19 13:56 . drwxr-xr-x 3 root root 60 Sep 19 15:52 .. crw-rw-r-- 1 root root 189, 256 Sep 18 13:19 001 crw-rw---- 1 root plugdev 189, 258 Sep 19 15:54 003 ```

kerem commented

Author

Owner

@Zixim commented on GitHub (Sep 19, 2025):

for me, the Coral tpu (usb), is working fine. With the exception that there might be a "visual" bug as shown here

@Zixim commented on GitHub (Sep 19, 2025): for me, the Coral tpu (usb), is working fine. With the exception that there might be a "visual" bug as shown [here](https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3306720462)

kerem commented

Author

Owner

@di5cord20 commented on GitHub (Sep 20, 2025):

I have frigate up and running now, but coral tpu isn't working yet.
frigate.detectors.plugins.edgetpu_tfl ERROR : Incorrect model used with EdgeTPU. Only .tflite models can be used with a Coral EdgeTPU.

from host - "lsusb"
Bus 002 Device 002: ID 18d1:9302 Google Inc.

from frigate lxc- "ls -la /dev/" (I don't have a bus directory)

Do I need to be manually passing through any devices? (i.e., /dev/bus/usb/002/ or /dev/dri/)

Other configs that may help for troubleshooting:

changed my lxc config to enable the USB Coral by adding the below lines:

lxc.mount.entry: /dev/bus/usb/002 dev/bus/usb/002 none bind,optional,create=dir

frigate config: (using default and haven't added any of my cameras yet)

detect:
  enabled: true
ffmpeg:
  hwaccel_args: auto
detectors:
  coral:
    type: edgetpu
    device: usb
  # detector01:
  #   type: openvino

When I update the above frigate config to use openvino instead:

--- Last 20 lines of silent log (/tmp/silent.2507140.log) ---
✔️ Storage 'local-lvm' supports 'rootdir'
✔️ Template storage 'local' supports 'vztmpl'

⠇ Starting LXC Container

Then it just hangs on Starting LXC Container forever.

Thanks again for the responses!
rhY

This is the error I get in the ProxMox GUI:
TASK ERROR: /dev/nvidia-caps is not a device

/edit

Tried again using advanced setup. Same result.

@crhy commented on GitHub (Oct 2, 2025): Thanks so much for the responses! Trying the development version now. So far I got this error during setup: Which GPU type to passthrough? (AMD NVIDIA): NVIDIA ✔️ NVIDIA GPU passthrough configured (6 devices) ⠸ Starting LXC Containervm 104 - unable to parse config: # NVIDIA GPU Passthrough (privileged) vm 104 - unable to parse config: lxc.cgroup2.devices.allow: c 195:* rwm vm 104 - unable to parse config: lxc.cgroup2.devices.allow: c 243:* rwm vm 104 - unable to parse config: lxc.mount.entry: /dev/nvidia0 dev/nvidia0 none bind,optional,create=file vm 104 - unable to parse config: lxc.mount.entry: /dev/nvidiactl dev/nvidiactl none bind,optional,create=file vm 104 - unable to parse config: lxc.mount.entry: /dev/nvidia-uvm dev/nvidia-uvm none bind,optional,create=file vm 104 - unable to parse config: lxc.mount.entry: /dev/nvidia-uvm-tools dev/nvidia-uvm-tools none bind,optional,create=file /dev/nvidia-caps is not a device [ERROR] in line 2423: exit code 255 (DPKG: Fatal internal error): while executing command pct start "$CTID" --- Last 20 lines of silent log (/tmp/silent.2507140.log) --- ✔️ Storage 'local-lvm' supports 'rootdir' ✔️ Template storage 'local' supports 'vztmpl' --------------------------------------------------- ⠇ Starting LXC Container _________________________________________ Then it just hangs on Starting LXC Container forever. Thanks again for the responses! rhY This is the error I get in the ProxMox GUI: TASK ERROR: /dev/nvidia-caps is not a device /edit Tried again using advanced setup. Same result.

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Oct 2, 2025):

Using the development version on Proxmox 8.4. Got the following:
💾 Disk Size: 50 GB
🧠 CPU Cores: 4
🛠️ RAM Size: 8192 MiB
🌉 Bridge: vmbr0
📡 IPv4: DHCP
📡 IPv6: SLAAC/AUTO
📡 APT-Cacher IP Address: Default
⚙️ Interface MTU Size: Default
🔍 DNS Search Domain: Host
📡 DNS Server IP Address: Host
🏷️ Vlan: Default
📡 Tags: community-script;nvr
🔑 Root SSH Access: yes
🗂️ Enable FUSE Support: no
🔍 Verbose Mode: yes
🚀 Creating a Frigate LXC using the above advanced settings
/dev/fd/63: line 1315: _load_vars_file_to_map: command not found
/dev/fd/63: line 1317: _load_vars_file_to_map: command not found
✔️ Storage ISO (Free: 36.0GB Used: 24.1GB) [Template]
✔️ Storage zVM (Free: 297.7GB Used: 580.6GB) [Container]
✔️ Cluster is quorate
✔️ Template debian-13-standard_13.1-2_amd64.tar.zst [local]
✔️ LXC Container 122 was successfully created.
✔️ USB passthrough configured
✔️ INTEL GPU passthrough configured (2 devices)
✔️ Started LXC Container
💡 Network reachable but gateway check failed
✔️ Device GIDs updated successfully
✔️ Customized LXC Container
✔️ Installed SSH keys into CT 122
✔️ Set up Container OS
✔️ Network Connected: 10.10.2.131 fd69:2160:eb9e:3144:be24:11ff:fe93:55bb
✔️ IPv4 Internet Connected
✖️ IPv6 Internet Not Connected
✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) git.community-scripts.org:(✔️ )
Hit:1 http://deb.debian.org/debian trixie InRelease
Hit:2 http://security.debian.org trixie-security InRelease
Hit:3 http://deb.debian.org/debian trixie-updates InRelease
Reading package lists... Done
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
Calculating upgrade... Done
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
✔️ Updated Container OS
Reading package lists... Done (Patience)
Building dependency tree... Done
Reading state information... Done
Note, selecting 'tcl' instead of 'tclsh'
Package libatlas-base-dev is not available, but is referred to by another package.
This may mean that the package is missing, has been obsoleted, or
is only available from another source

Package python3-distutils is not available, but is referred to by another package.
This may mean that the package is missing, has been obsoleted, or
is only available from another source

E: Package 'python3-distutils' has no installation candidate
E: Unable to locate package python3.11
E: Couldn't find any package by glob 'python3.11'
E: Couldn't find any package by regex 'python3.11'
E: Unable to locate package python3.11-dev
E: Couldn't find any package by glob 'python3.11-dev'
E: Couldn't find any package by regex 'python3.11-dev'
E: Package 'libatlas-base-dev' has no installation candidate

[ERROR] in line 18: exit code 18 (Unknown error): while executing command apt-get install -y {jq,wget,xz-utils,python3,python3-dev,python3-distutils,gcc,pkg-config,libhdf5-dev,unzip,build-essential,automake,libtool,ccache,libusb-1.0-0-dev,apt-transport-https,python3.11,python3.11-dev,cmake,git,libgtk-3-dev,libavcodec-dev,libavformat-dev,libswscale-dev,libv4l-dev,libxvidcore-dev,libx264-dev,libjpeg-dev,libpng-dev,libtiff-dev,gfortran,openexr,libatlas-base-dev,libssl-dev,libtbbmalloc2,libtbb-dev,libdc1394-dev,libopenexr-dev,libgstreamer-plugins-base1.0-dev,libgstreamer1.0-dev,tclsh,libopenblas-dev,liblapack-dev,make,moreutils}

After the command not found lines it asked me a 2nd time for the ISO and container locations then continued. But then failed on not finding python3

Tried on my Proxmox 9.0 system and got the same errors.

@atrozgrifo commented on GitHub (Oct 2, 2025): Using the development version on Proxmox 8.4. Got the following: 💾 Disk Size: 50 GB 🧠 CPU Cores: 4 🛠️ RAM Size: 8192 MiB 🌉 Bridge: vmbr0 📡 IPv4: DHCP 📡 IPv6: SLAAC/AUTO 📡 APT-Cacher IP Address: Default ⚙️ Interface MTU Size: Default 🔍 DNS Search Domain: Host 📡 DNS Server IP Address: Host 🏷️ Vlan: Default 📡 Tags: community-script;nvr 🔑 Root SSH Access: yes 🗂️ Enable FUSE Support: no 🔍 Verbose Mode: yes 🚀 Creating a Frigate LXC using the above advanced settings /dev/fd/63: line 1315: _load_vars_file_to_map: command not found /dev/fd/63: line 1317: _load_vars_file_to_map: command not found ✔️ Storage ISO (Free: 36.0GB Used: 24.1GB) [Template] ✔️ Storage zVM (Free: 297.7GB Used: 580.6GB) [Container] ✔️ Cluster is quorate ✔️ Template debian-13-standard_13.1-2_amd64.tar.zst [local] ✔️ LXC Container 122 was successfully created. ✔️ USB passthrough configured ✔️ INTEL GPU passthrough configured (2 devices) ✔️ Started LXC Container 💡 Network reachable but gateway check failed ✔️ Device GIDs updated successfully ✔️ Customized LXC Container ✔️ Installed SSH keys into CT 122 ✔️ Set up Container OS ✔️ Network Connected: 10.10.2.131 fd69:2160:eb9e:3144:be24:11ff:fe93:55bb ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) git.community-scripts.org:(✔️ ) Hit:1 http://deb.debian.org/debian trixie InRelease Hit:2 http://security.debian.org trixie-security InRelease Hit:3 http://deb.debian.org/debian trixie-updates InRelease Reading package lists... Done Reading package lists... Done Building dependency tree... Done Reading state information... Done Calculating upgrade... Done 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. ✔️ Updated Container OS Reading package lists... Done (Patience) Building dependency tree... Done Reading state information... Done Note, selecting 'tcl' instead of 'tclsh' Package libatlas-base-dev is not available, but is referred to by another package. This may mean that the package is missing, has been obsoleted, or is only available from another source Package python3-distutils is not available, but is referred to by another package. This may mean that the package is missing, has been obsoleted, or is only available from another source E: Package 'python3-distutils' has no installation candidate E: Unable to locate package python3.11 E: Couldn't find any package by glob 'python3.11' E: Couldn't find any package by regex 'python3.11' E: Unable to locate package python3.11-dev E: Couldn't find any package by glob 'python3.11-dev' E: Couldn't find any package by regex 'python3.11-dev' E: Package 'libatlas-base-dev' has no installation candidate [ERROR] in line 18: exit code 18 (Unknown error): while executing command apt-get install -y {jq,wget,xz-utils,python3,python3-dev,python3-distutils,gcc,pkg-config,libhdf5-dev,unzip,build-essential,automake,libtool,ccache,libusb-1.0-0-dev,apt-transport-https,python3.11,python3.11-dev,cmake,git,libgtk-3-dev,libavcodec-dev,libavformat-dev,libswscale-dev,libv4l-dev,libxvidcore-dev,libx264-dev,libjpeg-dev,libpng-dev,libtiff-dev,gfortran,openexr,libatlas-base-dev,libssl-dev,libtbbmalloc2,libtbb-dev,libdc1394-dev,libopenexr-dev,libgstreamer-plugins-base1.0-dev,libgstreamer1.0-dev,tclsh,libopenblas-dev,liblapack-dev,make,moreutils} After the command not found lines it asked me a 2nd time for the ISO and container locations then continued. But then failed on not finding python3 Tried on my Proxmox 9.0 system and got the same errors.

kerem commented

Author

Owner

@Zixim commented on GitHub (Oct 2, 2025):

@barryvdh I'm also noticing that the go2rtc config file isn't being populated as expected, but I understand you have it working.
Would you mind posting the full path to the file ?
I'm looking at /dev/shm/go2rtc.yaml , is that right ?

@Zixim commented on GitHub (Oct 2, 2025): @barryvdh I'm also noticing that the go2rtc config file isn't being populated as expected, but I understand you have it working. Would you mind posting the full path to the file ? I'm looking at `/dev/shm/go2rtc.yaml` , is that right ?

kerem commented

Author

Owner

@barryvdh commented on GitHub (Oct 2, 2025):

@barryvdh I'm also noticing that the go2rtc config file isn't being populated as expected, but I understand you have it working. Would you mind posting the full path to the file ? I'm looking at /dev/shm/go2rtc.yaml , is that right ?

Yes I think so, that is what it's using now. But not sure if there was and old file (with .yml instead of .yaml or something)

@barryvdh commented on GitHub (Oct 2, 2025): > [@barryvdh](https://github.com/barryvdh) I'm also noticing that the go2rtc config file isn't being populated as expected, but I understand you have it working. Would you mind posting the full path to the file ? I'm looking at `/dev/shm/go2rtc.yaml` , is that right ? Yes I think so, that is what it's using now. But not sure if there was and old file (with .yml instead of .yaml or something)

kerem commented

Author

Owner

@Zixim commented on GitHub (Oct 2, 2025):

@barryvdh

@barryvdh I'm also noticing that the go2rtc config file isn't being populated as expected, but I understand you have it working. Would you mind posting the full path to the file ? I'm looking at /dev/shm/go2rtc.yaml , is that right ?

Yes I think so, that is what it's using now. But not sure if there was and old file (with .yml instead of .yaml or something)

root@frigate:~# cat /dev/shm/go2rtc.yaml 
api:
  origin: '*'
hass:
  config: /homeassistant
log:
  format: text
webrtc:
  candidates:
  - stun:8555
ffmpeg:
  bin: /usr/lib/ffmpeg/7.0/bin/ffmpeg

And indeed there's no trace of the go2rtc: directives from inside of frigate's config.yml
What's in your go2rtc.yaml ?
Did you simply edit the actual file ?

@Zixim commented on GitHub (Oct 2, 2025): @barryvdh > > [@barryvdh](https://github.com/barryvdh) I'm also noticing that the go2rtc config file isn't being populated as expected, but I understand you have it working. Would you mind posting the full path to the file ? I'm looking at `/dev/shm/go2rtc.yaml` , is that right ? > > Yes I think so, that is what it's using now. But not sure if there was and old file (with .yml instead of .yaml or something) ```yaml root@frigate:~# cat /dev/shm/go2rtc.yaml api: origin: '*' hass: config: /homeassistant log: format: text webrtc: candidates: - stun:8555 ffmpeg: bin: /usr/lib/ffmpeg/7.0/bin/ffmpeg ``` And indeed there's no trace of the `go2rtc:` directives from inside of frigate's `config.yml` What's in your `go2rtc.yaml` ? Did you simply edit the actual file ?

kerem commented

Author

Owner

@barryvdh commented on GitHub (Oct 2, 2025):

The same but with the streams added. But it doesn't (always) save correctly.

I just go to port 1984 on Frigate to see go2rtc, and check the config / streams there.

@barryvdh commented on GitHub (Oct 2, 2025): The same but with the streams added. But it doesn't (always) save correctly. I just go to port 1984 on Frigate to see go2rtc, and check the config / streams there.

kerem commented

Author

Owner

@Zixim commented on GitHub (Oct 2, 2025):

@remz1337 confirmed : creating go2rtc streams via the Frigate config.yml file does not add the streams to go2rtc.
It does work if streams are added straight through the go2rtc web interface , as @barryvdh pointed out.
Even worse, you must also add the go2rtc: config in Frigate, if you don't then Frigate won't be able to do the camera name mapping. :/

@Zixim commented on GitHub (Oct 2, 2025): @remz1337 confirmed : creating go2rtc streams via the Frigate config.yml file does not add the streams to go2rtc. It does work if streams are added straight through the go2rtc web interface , as @barryvdh pointed out. Even worse, you must also add the go2rtc: config in Frigate, if you don't then Frigate won't be able to do the camera name mapping. :/

kerem commented

Author

Owner

@barryvdh commented on GitHub (Oct 2, 2025):

It also doesn't restart my container properly when restarting Frigate from the interface. I need to manually reboot the proxmox container for it to boot again, but not sure how/where to debug errors (getting 500 internal server error on nginx)

@barryvdh commented on GitHub (Oct 2, 2025): It also doesn't restart my container properly when restarting Frigate from the interface. I need to manually reboot the proxmox container for it to boot again, but not sure how/where to debug errors (getting 500 internal server error on nginx)

kerem commented

Author

Owner

@Zixim commented on GitHub (Oct 2, 2025):

It also doesn't restart my container properly when restarting Frigate from the interface. I need to manually reboot the proxmox container for it to boot again, but not sure how/where to debug errors (getting 500 internal server error on nginx)

For me, the webif functions "Save & Restart" (in config editor) as well as "Restart Frigate" (from the Settings menu) both properly restart Frigate + Go2rtc + Nginx. The LXC is not restarted, and isn't supposed to be - i think.

@Zixim commented on GitHub (Oct 2, 2025): > It also doesn't restart my container properly when restarting Frigate from the interface. I need to manually reboot the proxmox container for it to boot again, but not sure how/where to debug errors (getting 500 internal server error on nginx) For me, the webif functions "Save & Restart" (in config editor) as well as "Restart Frigate" (from the Settings menu) both properly restart Frigate + Go2rtc + Nginx. The LXC is not restarted, and isn't supposed to be - i think.

kerem commented

Author

Owner

@Honusnap commented on GitHub (Oct 2, 2025):

@remz1337 confirmed : creating go2rtc streams via the Frigate config.yml file does not add the streams to go2rtc. It does work if streams are added straight through the go2rtc web interface , as @barryvdh pointed out. Even worse, you must also add the go2rtc: config in Frigate, if you don't then Frigate won't be able to do the camera name mapping. :/

It was already the case with script 0.14 .... i rtan into this problem with it and installed go2rtc in a tandalone lxc .. like 1 year ago.

@Honusnap commented on GitHub (Oct 2, 2025): > [@remz1337](https://github.com/remz1337) confirmed : creating go2rtc streams via the Frigate config.yml file does not add the streams to go2rtc. It does work if streams are added straight through the go2rtc web interface , as [@barryvdh](https://github.com/barryvdh) pointed out. Even worse, you must also add the go2rtc: config in Frigate, if you don't then Frigate won't be able to do the camera name mapping. :/ It was already the case with script 0.14 .... i rtan into this problem with it and installed go2rtc in a tandalone lxc .. like 1 year ago.

kerem commented

Author

Owner

@Honusnap commented on GitHub (Oct 2, 2025):

It also doesn't restart my container properly when restarting Frigate from the interface. I need to manually reboot the proxmox container for it to boot again, but not sure how/where to debug errors (getting 500 internal server error on nginx)

For me, the webif functions "Save & Restart" (in config editor) as well as "Restart Frigate" (from the Settings menu) both properly restart Frigate + Go2rtc + Nginx. The LXC is not restarted, and isn't supposed to be - i think.

Agreed, lxc not supposed to be restarted,.

@Honusnap commented on GitHub (Oct 2, 2025): > > It also doesn't restart my container properly when restarting Frigate from the interface. I need to manually reboot the proxmox container for it to boot again, but not sure how/where to debug errors (getting 500 internal server error on nginx) > > For me, the webif functions "Save & Restart" (in config editor) as well as "Restart Frigate" (from the Settings menu) both properly restart Frigate + Go2rtc + Nginx. The LXC is not restarted, and isn't supposed to be - i think. Agreed, lxc not supposed to be restarted,.

kerem commented

Author

Owner

@Zixim commented on GitHub (Oct 3, 2025):

@remz1337 confirmed : creating go2rtc streams via the Frigate config.yml file does not add the streams to go2rtc. It does work if streams are added straight through the go2rtc web interface , as @barryvdh pointed out. Even worse, you must also add the go2rtc: config in Frigate, if you don't then Frigate won't be able to do the camera name mapping. :/

It was already the case with script 0.14 .... i rtan into this problem with it and installed go2rtc in a tandalone lxc .. like 1 year ago.

Can't confirm, my Frigate 0.14 is based on the Docker LXC + Frigate container.

@Zixim commented on GitHub (Oct 3, 2025): > > [@remz1337](https://github.com/remz1337) confirmed : creating go2rtc streams via the Frigate config.yml file does not add the streams to go2rtc. It does work if streams are added straight through the go2rtc web interface , as [@barryvdh](https://github.com/barryvdh) pointed out. Even worse, you must also add the go2rtc: config in Frigate, if you don't then Frigate won't be able to do the camera name mapping. :/ > > It was already the case with script 0.14 .... i rtan into this problem with it and installed go2rtc in a tandalone lxc .. like 1 year ago. Can't confirm, my Frigate 0.14 is based on the Docker LXC + Frigate container.

kerem commented

Author

Owner

@MichaelPankow commented on GitHub (Oct 4, 2025):

Now I was able to install frigate after removing OpenVino from installation script (cloning scripts, modify frigate-install.sh).
I am using a unpriviliged container, Coral is working after passthrough of /dev/apex_0. Go2rtc streams are also available after adding them directly into its conf file (/dev/shm/go2rtc.yaml).

@MichaelPankow commented on GitHub (Oct 4, 2025): Now I was able to install frigate after removing OpenVino from installation script (cloning scripts, modify frigate-install.sh). I am using a unpriviliged container, Coral is working after passthrough of /dev/apex_0. Go2rtc streams are also available after adding them directly into its conf file (/dev/shm/go2rtc.yaml).

kerem commented

Author

Owner

@poseidogithub commented on GitHub (Oct 4, 2025):

Is that a usb coral or internal?

I’m for me script works nice on a intel n100, but I can not make usb coral
work. (I’m sure it’s my fault, I have no clue!)

El El sáb, 4 oct 2025 a las 13:38, Michael_Pankow @.***>
escribió:

MichaelPankow left a comment (community-scripts/ProxmoxVE#6915)
https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3368150333

Now I was able to install frigate after removing OpenVino from
installation script (cloning scripts, modify frigate-install.sh).
I am using a unpriviliged container, Coral is working after passthrough of
/dev/apex_0. Go2rtc streams are also available after adding them directly
into its conf file (/dev/shm/go2rtc.yaml).

—
Reply to this email directly, view it on GitHub
https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3368150333,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/APUUUOURULTGSY73JJZ6FBD3V6WTTAVCNFSM6AAAAACEC6ZHIWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTGNRYGE2TAMZTGM
.
You are receiving this because you are subscribed to this thread.Message
ID: @.***>

@poseidogithub commented on GitHub (Oct 4, 2025): Is that a usb coral or internal? I’m for me script works nice on a intel n100, but I can not make usb coral work. (I’m sure it’s my fault, I have no clue!) El El sáb, 4 oct 2025 a las 13:38, Michael_Pankow ***@***.***> escribió: > *MichaelPankow* left a comment (community-scripts/ProxmoxVE#6915) > <https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3368150333> > > Now I was able to install frigate after removing OpenVino from > installation script (cloning scripts, modify frigate-install.sh). > I am using a unpriviliged container, Coral is working after passthrough of > /dev/apex_0. Go2rtc streams are also available after adding them directly > into its conf file (/dev/shm/go2rtc.yaml). > > — > Reply to this email directly, view it on GitHub > <https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3368150333>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/APUUUOURULTGSY73JJZ6FBD3V6WTTAVCNFSM6AAAAACEC6ZHIWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTGNRYGE2TAMZTGM> > . > You are receiving this because you are subscribed to this thread.Message > ID: ***@***.***> >

kerem commented

Author

Owner

@Zixim commented on GitHub (Oct 4, 2025):

/dev/apex_0 means it's a M.2 (Mini PCIe) coral TPU.
The USB coral is working just fine in Frigate for me.

@Zixim commented on GitHub (Oct 4, 2025): `/dev/apex_0` means it's a M.2 (Mini PCIe) coral TPU. The USB coral is working just fine in Frigate for me.

kerem commented

Author

Owner

@Zixim commented on GitHub (Oct 6, 2025):

Anyone here having success with face recognition ?
I'm experiencing that Frigate isn't restarting if face recognition is enabled, it just hangs.
Face recognition is working though.

@Zixim commented on GitHub (Oct 6, 2025): Anyone here having success with face recognition ? I'm experiencing that Frigate isn't restarting if face recognition is enabled, it just hangs. Face recognition is working though.

kerem commented

Author

Owner

@barryvdh commented on GitHub (Oct 6, 2025):

Anyone here having success with face recognition ? I'm experiencing that Frigate isn't restarting if face recognition is enabled, it just hangs. Face recognition is working though.

Face recognition is working fine. But restarting from the interface isn't (rebooting containers works, but doesn't seem specifically related to face recognition.

@barryvdh commented on GitHub (Oct 6, 2025): > Anyone here having success with face recognition ? I'm experiencing that Frigate isn't restarting if face recognition is enabled, it just hangs. Face recognition is working though. Face recognition is working fine. But restarting from the interface isn't (rebooting containers works, but doesn't seem specifically related to face recognition.

kerem commented

Author

Owner

@Zixim commented on GitHub (Oct 6, 2025):

For me the restart issue is 100% related to Face recognition.
If I stop frigate.service , remove the face_recognition config, and restart the service, then Frigate behaves as normal.
Add the face_recognition to the config, and restart Frigate. Frigate will start up & all works, but frigate will no longer restart from the webif. Stopping the service works, but it take a long time.

This is what happens when stopping frigate.service :

Oct 06 14:10:02 frigate systemd[1]: Stopping frigate.service - Frigate service...
Oct 06 14:11:33 frigate systemd[1]: frigate.service: State 'final-sigterm' timed out. Killing.
Oct 06 14:11:33 frigate systemd[1]: frigate.service: Killing process 1287057 (frigate.embeddi) with signal SIGKILL.
Oct 06 14:11:33 frigate systemd[1]: frigate.service: Killing process 1287079 (ZMQbg/Reaper) with signal SIGKILL.
Oct 06 14:11:33 frigate systemd[1]: frigate.service: Killing process 1287080 (ZMQbg/IO/0) with signal SIGKILL.
Oct 06 14:11:33 frigate systemd[1]: frigate.service: Failed with result 'timeout'.
Oct 06 14:11:33 frigate systemd[1]: Stopped frigate.service - Frigate service.

Of those 3 processes that are eventually killed, I think frigate.embeddings_manager is related to Face Recognition.

@remz1337 is frigate.embeddings_manager being sent a stop signal ?

@Zixim commented on GitHub (Oct 6, 2025): For me the restart issue is 100% related to Face recognition. If I stop frigate.service , remove the `face_recognition` config, and restart the service, then Frigate behaves as normal. Add the face_recognition to the config, and restart Frigate. Frigate will start up & all works, but frigate will no longer restart from the webif. Stopping the service works, but it take a long time. This is what happens when stopping frigate.service : ``` Oct 06 14:10:02 frigate systemd[1]: Stopping frigate.service - Frigate service... Oct 06 14:11:33 frigate systemd[1]: frigate.service: State 'final-sigterm' timed out. Killing. Oct 06 14:11:33 frigate systemd[1]: frigate.service: Killing process 1287057 (frigate.embeddi) with signal SIGKILL. Oct 06 14:11:33 frigate systemd[1]: frigate.service: Killing process 1287079 (ZMQbg/Reaper) with signal SIGKILL. Oct 06 14:11:33 frigate systemd[1]: frigate.service: Killing process 1287080 (ZMQbg/IO/0) with signal SIGKILL. Oct 06 14:11:33 frigate systemd[1]: frigate.service: Failed with result 'timeout'. Oct 06 14:11:33 frigate systemd[1]: Stopped frigate.service - Frigate service. ``` Of those 3 processes that are eventually killed, I think `frigate.embeddings_manager` is related to Face Recognition. @remz1337 is `frigate.embeddings_manager` being sent a stop signal ?

kerem commented

Author

Owner

@remz1337 commented on GitHub (Oct 6, 2025):

I don't know. Might be something to do with s6-overlay. From my understanding of Frigate (the official solution), all services and dependencies are managed by s6-overlay, which doesn't work in native linux environment (such as an LXC). To fix this, I had to port the service management to systemd, but I might have missed something.

Given the script has been reported to work for the most part by a few users, I will not be spending too much time on this.

Please remember, these are community scripts. Their goal is to help other users setup their services in Proxmox. This is not a paid product with defined features. Some services might not fully work or provide the full configurations/features available from the official solution. So, I encourage everyone contribute by opening PRs to improve/fix scripts that don't work to their liking.

@remz1337 commented on GitHub (Oct 6, 2025): I don't know. Might be something to do with s6-overlay. From my understanding of Frigate (the official solution), all services and dependencies are managed by s6-overlay, which doesn't work in native linux environment (such as an LXC). To fix this, I had to port the service management to systemd, but I might have missed something. Given the script has been reported to work for the most part by a few users, I will not be spending too much time on this. Please remember, these are community scripts. Their goal is to help other users setup their services in Proxmox. This is not a paid product with defined features. Some services might not fully work or provide the full configurations/features available from the official solution. So, I encourage everyone contribute by opening PRs to improve/fix scripts that don't work to their liking.

kerem commented

Author

Owner

@Zixim commented on GitHub (Oct 7, 2025):

I realize this is a community effort, and much respect to you and all others who are putting in the effort ! 🙏
Frigate just feels so tantalizingly close, and it's already better than the now-retired v0.14 !
Hopefully 0.16.x gets added to this repo some time soon ™️

@Zixim commented on GitHub (Oct 7, 2025): I realize this is a community effort, and much respect to you and all others who are putting in the effort ! 🙏 Frigate just feels so tantalizingly close, and it's already better than the now-retired v0.14 ! Hopefully 0.16.x gets added to this repo some time soon ™️

kerem commented

Author

Owner

@samwathegreat commented on GitHub (Oct 10, 2025):

Amazing work -- Really excited to see a release soon!! Out of curiosity, do we think it will still be bookworm-based, or trixie?

@samwathegreat commented on GitHub (Oct 10, 2025): Amazing work -- Really excited to see a release soon!! Out of curiosity, do we think it will still be bookworm-based, or trixie?

kerem commented

Author

Owner

@albotroz commented on GitHub (Oct 12, 2025):

Why was version 0.14 removed from the community scripts mainpage?
The version still works for me—without any problems.

Were there any serious issues that led to it being removed?

@albotroz commented on GitHub (Oct 12, 2025): Why was version 0.14 removed from the community scripts mainpage? The version still works for me—without any problems. Were there any serious issues that led to it being removed?

kerem commented

Author

Owner

@jhogendorn commented on GitHub (Oct 12, 2025):

I'm running into

Setting up ssh (1:9.2p1-2+deb12u7) ...
Processing triggers for systemd (252.39-1~deb12u1) ...
Processing triggers for man-db (2.11.2-2) ...
Processing triggers for dbus (1.14.10-1~deb12u1) ...
Processing triggers for debianutils (5.7-0.5~deb12u1) ...
Processing triggers for mailcap (3.70+nmu1) ...
Processing triggers for libc-bin (2.36-9+deb12u13) ...
Processing triggers for ca-certificates (20230311+deb12u1) ...
Updating certificates in /etc/ssl/certs...
0 added, 0 removed; done.
Running hooks in /etc/ca-certificates/update.d...
done.
  ✔️   Updated Container OS
  ✔️   Set up Container
  ✔️   Installed APT proxy client
  ✔️   Installed sudo
  ✔️   Added SSH user jhogendorn (sudo)
  ✔️   Mounted shared directory
  ⏳   Rebooting LXCrun_buffer: 571 Script exited with status 2
lxc_init: 845 Failed to run lxc.hook.pre-start for container "700"
__lxc_start: 2034 Failed to initialize container "700"
startup for container '700' failed
/dev/fd/63: line 103: pop_var_context: head of shell_variables not a function context

thats using: bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"

my config:

+ cat /opt/community-scripts/frigate.conf
# frigate Configuration File
# Generated on Sun Oct 12 12:51:39 PM AEST 2025

CT_TYPE="1"
DISK_SIZE="50"
CORE_COUNT="4"
RAM_SIZE="16384"
VERBOSE="yes"
PW="pwd"
#CT_ID=101
HN="frigate"
BRG="vmbr0"
NET="dhcp"
IPV6_METHOD="auto"

# Set this only if using "IPV6_METHOD=static"
#IPV6STATIC="fd00::1234/64"

GATE="none"
APT_CACHER_IP="none"
MTU="1500"
SD="mydomain.com"
NS="none"
MAC="none"
VLAN="none"
SSH="no"
SSH_AUTHORIZED_KEY=""
TAGS="remz1337;nvr"
ENABLE_FUSE="yes"
ENABLE_TUN="no"

there were other config options it prompted for but apparently those didnt get pushed to the config.

I had a dig through the scripts but theres a level of indirection there and i cant figure out where the actual code is.

any suggestions?

I am trying to do an unprivileged install, if it matters. i dont have a gpu but i do have an m.2 coral tpu. reason for unprivileged is that i want to mount in my media dir as an nfs share, and all my auth management is nfs+kerberos+ldap and because of the factors of lxc containers its been proving extremely tricky to get things to act or map to the correct service account for permissions...

@jhogendorn commented on GitHub (Oct 12, 2025): I'm running into ``` Setting up ssh (1:9.2p1-2+deb12u7) ... Processing triggers for systemd (252.39-1~deb12u1) ... Processing triggers for man-db (2.11.2-2) ... Processing triggers for dbus (1.14.10-1~deb12u1) ... Processing triggers for debianutils (5.7-0.5~deb12u1) ... Processing triggers for mailcap (3.70+nmu1) ... Processing triggers for libc-bin (2.36-9+deb12u13) ... Processing triggers for ca-certificates (20230311+deb12u1) ... Updating certificates in /etc/ssl/certs... 0 added, 0 removed; done. Running hooks in /etc/ca-certificates/update.d... done. ✔️ Updated Container OS ✔️ Set up Container ✔️ Installed APT proxy client ✔️ Installed sudo ✔️ Added SSH user jhogendorn (sudo) ✔️ Mounted shared directory ⏳ Rebooting LXCrun_buffer: 571 Script exited with status 2 lxc_init: 845 Failed to run lxc.hook.pre-start for container "700" __lxc_start: 2034 Failed to initialize container "700" startup for container '700' failed /dev/fd/63: line 103: pop_var_context: head of shell_variables not a function context ``` thats using: `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"` my config: ``` + cat /opt/community-scripts/frigate.conf # frigate Configuration File # Generated on Sun Oct 12 12:51:39 PM AEST 2025 CT_TYPE="1" DISK_SIZE="50" CORE_COUNT="4" RAM_SIZE="16384" VERBOSE="yes" PW="pwd" #CT_ID=101 HN="frigate" BRG="vmbr0" NET="dhcp" IPV6_METHOD="auto" # Set this only if using "IPV6_METHOD=static" #IPV6STATIC="fd00::1234/64" GATE="none" APT_CACHER_IP="none" MTU="1500" SD="mydomain.com" NS="none" MAC="none" VLAN="none" SSH="no" SSH_AUTHORIZED_KEY="" TAGS="remz1337;nvr" ENABLE_FUSE="yes" ENABLE_TUN="no" ``` there were other config options it prompted for but apparently those didnt get pushed to the config. I had a dig through the scripts but theres a level of indirection there and i cant figure out where the actual code is. any suggestions? I am trying to do an unprivileged install, if it matters. i dont have a gpu but i do have an m.2 coral tpu. reason for unprivileged is that i want to mount in my media dir as an nfs share, and all my auth management is nfs+kerberos+ldap and because of the factors of lxc containers its been proving extremely tricky to get things to act or map to the correct service account for permissions...

kerem commented

Author

Owner

@remz1337 commented on GitHub (Oct 12, 2025):

@jhogendorn my advice, start with a basic default install, then move on from there (changing 1 thing at a time). Looks like you changed quite a few configs.

@remz1337 commented on GitHub (Oct 12, 2025): @jhogendorn my advice, start with a basic default install, then move on from there (changing 1 thing at a time). Looks like you changed quite a few configs.

kerem commented

Author

Owner

@wlad1164 commented on GitHub (Oct 12, 2025):

    ______     _             __     
   / ____/____(_)___ _____ _/ /____ 
  / /_  / ___/ / __ `/ __ `/ __/ _ \
 / __/ / /  / / /_/ / /_/ / /_/  __/
/_/   /_/  /_/\__, /\__,_/\__/\___/ 
             /____/                 
  🧩  Using Advanced Settings on node pve
  🖥️  Operating System: debian
  🌟  Version: 12
  📦  Container Type: Privileged
  🔐  Root Password: Automatic Login
  🆔  Container ID: 102
  🏠  Hostname: frigate16
  💾  Disk Size: 50 GB
  🧠  CPU Cores: 4
  🛠️  RAM Size: 4096 MiB
  🌉  Bridge: vmbr0
  📡  IPv4: DHCP
  📡  IPv6: SLAAC/AUTO
  📡  APT-Cacher IP Address: Default
  ⚙️  Interface MTU Size: Default
  🔍  DNS Search Domain: Host
  📡  DNS Server IP Address: Host
Enable Firewall: 0
  🏷️  Vlan: Default
  📡  Tags: remz1337;nvr
  🔑  Root SSH Access: no
  🗂️  Enable FUSE Support: no
   ✖️   Missing proxmox-helper-scripts environment variables: SSH_USER, SSH_PASSWORD
Setting Proxmox-Helper-Scripts Envrionment Variable SSH_USER: admin
Setting Proxmox-Helper-Scripts Envrionment Variable SSH_PASSWORD: ******
Add common sudo user with SSH access: yes
Enable Shared Mount: no
Configure Postfix Satellite: no
Nvidia GPU Passthrough: no
  🔍  Verbose Mode: no
  🚀  Creating a Frigate LXC using the above advanced settings
  ✔️   Storage local (Free: 79.1GB  Used: 8.0GB) [Template]
  ✔️   Storage local-lvm (Free: 312.4GB  Used: 28.4GB) [Container]
  ✔️   Template debian-12-standard_12.12-1_amd64.tar.zst [local]
  ✔️   LXC Container 102 was successfully created.

 ⚙️  Configuring VAAPI passthrough for LXC container
 ℹ️  VAAPI enables GPU hardware acceleration (e.g., for video transcoding in Jellyfin or Plex).

➤ Automatically mount all available VAAPI devices? [Y/n]: Y
  ✔️   Started LXC Container
   💡   No network in LXC yet (try 1/10) – waiting...
  ✔️   Network in LXC is reachable (ping)
  ✔️   Customized LXC Container
Customizing LXC creation
  ✔️   Rebooted LXC
  ✔️   Set up Container OS
  ✔️   Network Connected: 192.168.10.231 fd59:901c:9321:0:be24:11ff:fe3b:9095 
  ✔️   IPv4 Internet Connected
   ✖️   IPv6 Internet Not Connected
  ✔️   Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
  ✔️   Updated Container OS
  ✔️   Set up Container
  ✔️   Installed APT proxy client
  ✔️   Installed sudo
  ✔️   Added SSH user admin (sudo)
  ✔️   Post install script completed.
  ✔️   Set up Container OS
  ✔️   Network Connected: 192.168.10.231 fd59:901c:9321:0:be24:11ff:fe3b:9095 
  ✔️   IPv4 Internet Connected
   ✖️   IPv6 Internet Not Connected
  ✔️   Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
  ✔️   Updated Container OS
  ✔️   Installed Dependencies
  ✔️   Set Up Hardware Acceleration
  ✔️   Setup environment
  ✔️   Deployed: frigate (0.16.1)
  ✔️   Downloaded Frigate source
  ✔️   Built Nginx
  ✔️   Built SQLite
  ✔️   Deployed: go2rtc (1.9.10)
  ✔️   Installed go2rtc
  ✔️   Installed Tempio
  ✔️   Built libUSB
  ✔️   Installed Pip
  ✔️   Installed Frigate Dependencies
  ✔️   Built pysqlite3
  ✔️   Setup Node.js 22
  ✔️   Installed Node.js modules: yarn
  ✔️   Installed NodeJS
  ✔️   Downloaded Coral TPU Model
  ✔️   Downloaded CPU Model
  ✔️   Built Audio Models
⠼ Building HailoRT
[ERROR] in line 159: exit code 0: while executing command "$@" > /dev/null 2>&1
/dev/fd/63: line 173: pop_var_context: head of shell_variables not a function context
/dev/fd/63: line 288: local: can only be used in a function
/dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context

I am facing this error
Proxmox 8.3.5
N100

@wlad1164 commented on GitHub (Oct 12, 2025): ``` ______ _ __ / ____/____(_)___ _____ _/ /____ / /_ / ___/ / __ `/ __ `/ __/ _ \ / __/ / / / / /_/ / /_/ / /_/ __/ /_/ /_/ /_/\__, /\__,_/\__/\___/ /____/ 🧩 Using Advanced Settings on node pve 🖥️ Operating System: debian 🌟 Version: 12 📦 Container Type: Privileged 🔐 Root Password: Automatic Login 🆔 Container ID: 102 🏠 Hostname: frigate16 💾 Disk Size: 50 GB 🧠 CPU Cores: 4 🛠️ RAM Size: 4096 MiB 🌉 Bridge: vmbr0 📡 IPv4: DHCP 📡 IPv6: SLAAC/AUTO 📡 APT-Cacher IP Address: Default ⚙️ Interface MTU Size: Default 🔍 DNS Search Domain: Host 📡 DNS Server IP Address: Host Enable Firewall: 0 🏷️ Vlan: Default 📡 Tags: remz1337;nvr 🔑 Root SSH Access: no 🗂️ Enable FUSE Support: no ✖️ Missing proxmox-helper-scripts environment variables: SSH_USER, SSH_PASSWORD Setting Proxmox-Helper-Scripts Envrionment Variable SSH_USER: admin Setting Proxmox-Helper-Scripts Envrionment Variable SSH_PASSWORD: ****** Add common sudo user with SSH access: yes Enable Shared Mount: no Configure Postfix Satellite: no Nvidia GPU Passthrough: no 🔍 Verbose Mode: no 🚀 Creating a Frigate LXC using the above advanced settings ✔️ Storage local (Free: 79.1GB Used: 8.0GB) [Template] ✔️ Storage local-lvm (Free: 312.4GB Used: 28.4GB) [Container] ✔️ Template debian-12-standard_12.12-1_amd64.tar.zst [local] ✔️ LXC Container 102 was successfully created. ⚙️ Configuring VAAPI passthrough for LXC container ℹ️ VAAPI enables GPU hardware acceleration (e.g., for video transcoding in Jellyfin or Plex). ➤ Automatically mount all available VAAPI devices? [Y/n]: Y ✔️ Started LXC Container 💡 No network in LXC yet (try 1/10) – waiting... ✔️ Network in LXC is reachable (ping) ✔️ Customized LXC Container Customizing LXC creation ✔️ Rebooted LXC ✔️ Set up Container OS ✔️ Network Connected: 192.168.10.231 fd59:901c:9321:0:be24:11ff:fe3b:9095 ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) ✔️ Updated Container OS ✔️ Set up Container ✔️ Installed APT proxy client ✔️ Installed sudo ✔️ Added SSH user admin (sudo) ✔️ Post install script completed. ✔️ Set up Container OS ✔️ Network Connected: 192.168.10.231 fd59:901c:9321:0:be24:11ff:fe3b:9095 ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) ✔️ Updated Container OS ✔️ Installed Dependencies ✔️ Set Up Hardware Acceleration ✔️ Setup environment ✔️ Deployed: frigate (0.16.1) ✔️ Downloaded Frigate source ✔️ Built Nginx ✔️ Built SQLite ✔️ Deployed: go2rtc (1.9.10) ✔️ Installed go2rtc ✔️ Installed Tempio ✔️ Built libUSB ✔️ Installed Pip ✔️ Installed Frigate Dependencies ✔️ Built pysqlite3 ✔️ Setup Node.js 22 ✔️ Installed Node.js modules: yarn ✔️ Installed NodeJS ✔️ Downloaded Coral TPU Model ✔️ Downloaded CPU Model ✔️ Built Audio Models ⠼ Building HailoRT [ERROR] in line 159: exit code 0: while executing command "$@" > /dev/null 2>&1 /dev/fd/63: line 173: pop_var_context: head of shell_variables not a function context /dev/fd/63: line 288: local: can only be used in a function /dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context ``` I am facing this error Proxmox 8.3.5 N100

kerem commented

Author

Owner

@xenoncorp2k commented on GitHub (Oct 13, 2025):

Idem:

`   ______     _             __     
   / ____/____(_)___ _____ _/ /____ 
  / /_  / ___/ / __ `/ __ `/ __/ _ \
 / __/ / /  / / /_/ / /_/ / /_/  __/
/_/   /_/  /_/\__, /\__,_/\__/\___/ 
             /____/                 
  🧩  Using Advanced Settings on node proxmox1
  🖥️  Operating System: debian
  🌟  Version: 12
  📦  Container Type: Privileged
  🔐  Root Password: ********
  🆔  Container ID: 401
  🏠  Hostname: frigate16
  💾  Disk Size: 50 GB
  🧠  CPU Cores: 4
  🛠️  RAM Size: 8192 MiB
  🌉  Bridge: vmbr0
  📡  IPv4 Address: 192.168.1.42/24
  🌐  Gateway IP Address: 192.168.1.1
  📡  IPv6: Disabled
  📡  APT-Cacher IP Address: Default
  ⚙️  Interface MTU Size: Default
  🔍  DNS Search Domain: Host
  📡  DNS Server IP Address: Host
Enable Firewall: 0
  🏷️  Vlan: Default
  📡  Tags: remz1337;nvr
  🔑  Root SSH Access: no
  🗂️  Enable FUSE Support: no
Add common sudo user with SSH access: no
Enable Shared Mount: no
Configure Postfix Satellite: no
Nvidia GPU Passthrough: no
  🔍  Verbose Mode: no
  🚀  Creating a Frigate LXC using the above advanced settings
  ✔️   Storage local (Free: 70.1GB  Used: 42.7GB) [Template]
  ✔️   Storage local-lvm (Free: 234.1GB  Used: 61.3GB) [Container]
  ✔️   Cluster is quorate
  ✔️   Template debian-12-standard_12.12-1_amd64.tar.zst [local]
  ✔️   LXC Container 401 was successfully created.

 ⚙️  Configuring VAAPI passthrough for LXC container
 ℹ️  VAAPI enables GPU hardware acceleration (e.g., for video transcoding in Jellyfin or Plex).

➤ Automatically mount all available VAAPI devices? [Y/n]: Y
  ✔️   Started LXC Container
  ✔️   Network in LXC is reachable (ping)
  ✔️   Customized LXC Container
Customizing LXC creation
  ✔️   Rebooted LXC
  ✔️   Set up Container OS
  ✔️   Network Connected: 192.168.1.42 
  ✔️   IPv4 Internet Connected
   ✖️   IPv6 Internet Not Connected
  ✔️   Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
  ✔️   Updated Container OS
  ✔️   Set up Container
  ✔️   Installed APT proxy client
  ✔️   Installed sudo
  ✔️   Post install script completed.
  ✔️   Set up Container OS
  ✔️   Network Connected: 192.168.1.42 
  ✔️   IPv4 Internet Connected
   ✖️   IPv6 Internet Not Connected
  ✔️   Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
  ✔️   Updated Container OS
  ✔️   Installed Dependencies
  ✔️   Set Up Hardware Acceleration
  ✔️   Setup environment
  ✔️   Deployed: frigate (0.16.1)
  ✔️   Downloaded Frigate source
  ✔️   Built Nginx
  ✔️   Built SQLite
  ✔️   Deployed: go2rtc (1.9.10)
  ✔️   Installed go2rtc
  ✔️   Installed Tempio
  ✔️   Built libUSB
  ✔️   Installed Pip
  ✔️   Installed Frigate Dependencies
  ✔️   Built pysqlite3
  ✔️   Setup Node.js 22
  ✔️   Installed Node.js modules: yarn
  ✔️   Installed NodeJS
  ✔️   Downloaded Coral TPU Model
  ✔️   Downloaded CPU Model
  ✔️   Built Audio Models
⠸ Building HailoRT
[ERROR] in line 159: exit code 0: while executing command "$@" > /dev/null 2>&1
/dev/fd/63: line 173: pop_var_context: head of shell_variables not a function context
/dev/fd/63: line 288: local: can only be used in a function
/dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context

Idem error
Proxmox 9.0.10
i7-7700

@xenoncorp2k commented on GitHub (Oct 13, 2025): Idem: ``` ` ______ _ __ / ____/____(_)___ _____ _/ /____ / /_ / ___/ / __ `/ __ `/ __/ _ \ / __/ / / / / /_/ / /_/ / /_/ __/ /_/ /_/ /_/\__, /\__,_/\__/\___/ /____/ 🧩 Using Advanced Settings on node proxmox1 🖥️ Operating System: debian 🌟 Version: 12 📦 Container Type: Privileged 🔐 Root Password: ******** 🆔 Container ID: 401 🏠 Hostname: frigate16 💾 Disk Size: 50 GB 🧠 CPU Cores: 4 🛠️ RAM Size: 8192 MiB 🌉 Bridge: vmbr0 📡 IPv4 Address: 192.168.1.42/24 🌐 Gateway IP Address: 192.168.1.1 📡 IPv6: Disabled 📡 APT-Cacher IP Address: Default ⚙️ Interface MTU Size: Default 🔍 DNS Search Domain: Host 📡 DNS Server IP Address: Host Enable Firewall: 0 🏷️ Vlan: Default 📡 Tags: remz1337;nvr 🔑 Root SSH Access: no 🗂️ Enable FUSE Support: no Add common sudo user with SSH access: no Enable Shared Mount: no Configure Postfix Satellite: no Nvidia GPU Passthrough: no 🔍 Verbose Mode: no 🚀 Creating a Frigate LXC using the above advanced settings ✔️ Storage local (Free: 70.1GB Used: 42.7GB) [Template] ✔️ Storage local-lvm (Free: 234.1GB Used: 61.3GB) [Container] ✔️ Cluster is quorate ✔️ Template debian-12-standard_12.12-1_amd64.tar.zst [local] ✔️ LXC Container 401 was successfully created. ⚙️ Configuring VAAPI passthrough for LXC container ℹ️ VAAPI enables GPU hardware acceleration (e.g., for video transcoding in Jellyfin or Plex). ➤ Automatically mount all available VAAPI devices? [Y/n]: Y ✔️ Started LXC Container ✔️ Network in LXC is reachable (ping) ✔️ Customized LXC Container Customizing LXC creation ✔️ Rebooted LXC ✔️ Set up Container OS ✔️ Network Connected: 192.168.1.42 ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) ✔️ Updated Container OS ✔️ Set up Container ✔️ Installed APT proxy client ✔️ Installed sudo ✔️ Post install script completed. ✔️ Set up Container OS ✔️ Network Connected: 192.168.1.42 ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) ✔️ Updated Container OS ✔️ Installed Dependencies ✔️ Set Up Hardware Acceleration ✔️ Setup environment ✔️ Deployed: frigate (0.16.1) ✔️ Downloaded Frigate source ✔️ Built Nginx ✔️ Built SQLite ✔️ Deployed: go2rtc (1.9.10) ✔️ Installed go2rtc ✔️ Installed Tempio ✔️ Built libUSB ✔️ Installed Pip ✔️ Installed Frigate Dependencies ✔️ Built pysqlite3 ✔️ Setup Node.js 22 ✔️ Installed Node.js modules: yarn ✔️ Installed NodeJS ✔️ Downloaded Coral TPU Model ✔️ Downloaded CPU Model ✔️ Built Audio Models ⠸ Building HailoRT [ERROR] in line 159: exit code 0: while executing command "$@" > /dev/null 2>&1 /dev/fd/63: line 173: pop_var_context: head of shell_variables not a function context /dev/fd/63: line 288: local: can only be used in a function /dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context ``` Idem error Proxmox 9.0.10 i7-7700

kerem commented

Author

Owner

@AsmoOne commented on GitHub (Oct 13, 2025):

@xenoncorp2k Hi! i have same problem during HailoRR installation. My cpu is I5-6500T and proxmox 9.0.5

@AsmoOne commented on GitHub (Oct 13, 2025): @xenoncorp2k Hi! i have same problem during HailoRR installation. My cpu is I5-6500T and proxmox 9.0.5

kerem commented

Author

Owner

@remz1337 commented on GitHub (Oct 15, 2025):

@wlad1164 @xenoncorp2k can you run again with verbose output and paste the log back here again? thanks

@remz1337 commented on GitHub (Oct 15, 2025): @wlad1164 @xenoncorp2k can you run again with verbose output and paste the log back here again? thanks

kerem commented

Author

Owner

@wlad1164 commented on GitHub (Oct 15, 2025):

@remz1337
______ _ __
/ /()__ _____ / /___
/ /_ / / / __ / __ / __/ _
/ __/ / / / / // / // / // /
// // //_, /_,/__/_/
/____/
🧩 Using Advanced Settings on node pve
🖥️ Operating System: debian
🌟 Version: 12
📦 Container Type: Privileged
🔐 Root Password: Automatic Login
🆔 Container ID: 102
🏠 Hostname: frigate16
💾 Disk Size: 50 GB
🧠 CPU Cores: 4
🛠️ RAM Size: 4096 MiB
🌉 Bridge: vmbr0
📡 IPv4: DHCP
📡 IPv6: SLAAC/AUTO
📡 APT-Cacher IP Address: Default
⚙️ Interface MTU Size: Default
🔍 DNS Search Domain: Host
📡 DNS Server IP Address: Host
Enable Firewall: 0
🏷️ Vlan: Default
📡 Tags: remz1337;nvr
🔑 Root SSH Access: no
🗂️ Enable FUSE Support: no
Add common sudo user with SSH access: no
Enable Shared Mount: no
Configure Postfix Satellite: no
Nvidia GPU Passthrough: no
🔍 Verbose Mode: no
🚀 Creating a Frigate LXC using the above advanced settings
✔️ Storage local (Free: 79.1GB Used: 8.0GB) [Template]
✔️ Storage local-lvm (Free: 313.1GB Used: 27.7GB) [Container]
✔️ Template debian-12-standard_12.12-1_amd64.tar.zst [local]
✔️ LXC Container 102 was successfully created.

⚙️ Configuring VAAPI passthrough for LXC container
ℹ️ VAAPI enables GPU hardware acceleration (e.g., for video transcoding in Jellyfin or Plex).

➤ Automatically mount all available VAAPI devices? [Y/n]: Y
✔️ Started LXC Container
💡 No network in LXC yet (try 1/10) – waiting...
✔️ Network in LXC is reachable (ping)
✔️ Customized LXC Container
Customizing LXC creation
✔️ Rebooted LXC
✔️ Set up Container OS
✔️ Network Connected: 192.168.10.187 fd59:901c:9321:0:be24:11ff:fe84:5b89
✔️ IPv4 Internet Connected
✖️ IPv6 Internet Not Connected
✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
✔️ Updated Container OS
✔️ Set up Container
✔️ Installed APT proxy client
✔️ Installed sudo
✔️ Post install script completed.
✔️ Set up Container OS
✔️ Network Connected: 192.168.10.187 fd59:901c:9321:0:be24:11ff:fe84:5b89
✔️ IPv4 Internet Connected
✖️ IPv6 Internet Not Connected
✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
✔️ Updated Container OS
✔️ Installed Dependencies
✔️ Set Up Hardware Acceleration
✔️ Setup environment
✔️ Deployed: frigate (0.16.2)
✔️ Downloaded Frigate source
✔️ Built Nginx
✔️ Built SQLite
✔️ Deployed: go2rtc (1.9.10)
✔️ Installed go2rtc
✔️ Installed Tempio
✔️ Built libUSB
✔️ Installed Pip
✔️ Installed Frigate Dependencies
✔️ Built pysqlite3
✔️ Setup Node.js 22
✔️ Installed Node.js modules: yarn
✔️ Installed NodeJS
✔️ Downloaded Coral TPU Model
✔️ Downloaded CPU Model
✔️ Built Audio Models
⠦ Building HailoRT
[ERROR] in line 159: exit code 0: while executing command "$@" > /dev/null 2>&1
/dev/fd/63: line 173: pop_var_context: head of shell_variables not a function context
/dev/fd/63: line 288: local: can only be used in a function
/dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context
root@pve:~#

I use
bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"
Proxmox 8.3.5
N100

@wlad1164 commented on GitHub (Oct 15, 2025): @remz1337 ______ _ __ / ____/____(_)___ _____ _/ /____ / /_ / ___/ / __ `/ __ `/ __/ _ \ / __/ / / / / /_/ / /_/ / /_/ __/ /_/ /_/ /_/\__, /\__,_/\__/\___/ /____/ 🧩 Using Advanced Settings on node pve 🖥️ Operating System: debian 🌟 Version: 12 📦 Container Type: Privileged 🔐 Root Password: Automatic Login 🆔 Container ID: 102 🏠 Hostname: frigate16 💾 Disk Size: 50 GB 🧠 CPU Cores: 4 🛠️ RAM Size: 4096 MiB 🌉 Bridge: vmbr0 📡 IPv4: DHCP 📡 IPv6: SLAAC/AUTO 📡 APT-Cacher IP Address: Default ⚙️ Interface MTU Size: Default 🔍 DNS Search Domain: Host 📡 DNS Server IP Address: Host Enable Firewall: 0 🏷️ Vlan: Default 📡 Tags: remz1337;nvr 🔑 Root SSH Access: no 🗂️ Enable FUSE Support: no Add common sudo user with SSH access: no Enable Shared Mount: no Configure Postfix Satellite: no Nvidia GPU Passthrough: no 🔍 Verbose Mode: no 🚀 Creating a Frigate LXC using the above advanced settings ✔️ Storage local (Free: 79.1GB Used: 8.0GB) [Template] ✔️ Storage local-lvm (Free: 313.1GB Used: 27.7GB) [Container] ✔️ Template debian-12-standard_12.12-1_amd64.tar.zst [local] ✔️ LXC Container 102 was successfully created. ⚙️ Configuring VAAPI passthrough for LXC container ℹ️ VAAPI enables GPU hardware acceleration (e.g., for video transcoding in Jellyfin or Plex). ➤ Automatically mount all available VAAPI devices? [Y/n]: Y ✔️ Started LXC Container 💡 No network in LXC yet (try 1/10) – waiting... ✔️ Network in LXC is reachable (ping) ✔️ Customized LXC Container Customizing LXC creation ✔️ Rebooted LXC ✔️ Set up Container OS ✔️ Network Connected: 192.168.10.187 fd59:901c:9321:0:be24:11ff:fe84:5b89 ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) ✔️ Updated Container OS ✔️ Set up Container ✔️ Installed APT proxy client ✔️ Installed sudo ✔️ Post install script completed. ✔️ Set up Container OS ✔️ Network Connected: 192.168.10.187 fd59:901c:9321:0:be24:11ff:fe84:5b89 ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) ✔️ Updated Container OS ✔️ Installed Dependencies ✔️ Set Up Hardware Acceleration ✔️ Setup environment ✔️ Deployed: frigate (0.16.2) ✔️ Downloaded Frigate source ✔️ Built Nginx ✔️ Built SQLite ✔️ Deployed: go2rtc (1.9.10) ✔️ Installed go2rtc ✔️ Installed Tempio ✔️ Built libUSB ✔️ Installed Pip ✔️ Installed Frigate Dependencies ✔️ Built pysqlite3 ✔️ Setup Node.js 22 ✔️ Installed Node.js modules: yarn ✔️ Installed NodeJS ✔️ Downloaded Coral TPU Model ✔️ Downloaded CPU Model ✔️ Built Audio Models ⠦ Building HailoRT [ERROR] in line 159: exit code 0: while executing command "$@" > /dev/null 2>&1 /dev/fd/63: line 173: pop_var_context: head of shell_variables not a function context /dev/fd/63: line 288: local: can only be used in a function /dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context root@pve:~# I use `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"` Proxmox 8.3.5 N100

kerem commented

Author

Owner

@remz1337 commented on GitHub (Oct 15, 2025):

@wlad1164 @xenoncorp2k can you run again with verbose output and paste the log back here again? thanks

I need verbose output

@remz1337 commented on GitHub (Oct 15, 2025): > [@wlad1164](https://github.com/wlad1164) [@xenoncorp2k](https://github.com/xenoncorp2k) can you run again with verbose output and paste the log back here again? thanks I need verbose output

kerem commented

Author

Owner

@samwathegreat commented on GitHub (Oct 15, 2025):

@wlad1164 @xenoncorp2k can you run again with verbose output and paste the log back here again? thanks

I need verbose output

Hmm. When I run verbose output on a script, it looks just like what @wlad1164 pasted. How do they get you what you need?

@samwathegreat commented on GitHub (Oct 15, 2025): > > [@wlad1164](https://github.com/wlad1164) [@xenoncorp2k](https://github.com/xenoncorp2k) can you run again with verbose output and paste the log back here again? thanks > > I need verbose output Hmm. When I run verbose output on a script, it looks just like what @wlad1164 pasted. How do they get you what you need?

kerem commented

Author

Owner

@wlad1164 commented on GitHub (Oct 15, 2025):

@wlad1164 @xenoncorp2k can you run again with verbose output and paste the log back here again? thanks

✔️ Installed Dependencies
Reading package lists... Doneeleration
Building dependency tree... Done
Reading state information... Done
va-driver-all is already the newest version (2.17.0-1).
va-driver-all set to manually installed.
ocl-icd-libopencl1 is already the newest version (2.3.1-1).
ocl-icd-libopencl1 set to manually installed.
The following additional packages will be installed:
libclang-cpp14 libigc1 libigdfcl1 libllvm14 libllvmspirvlib14 libopencl-clang14 libva-wayland2 libxv1
The following NEW packages will be installed:
intel-gpu-tools intel-opencl-icd libclang-cpp14 libigc1 libigdfcl1 libllvm14 libllvmspirvlib14 libopencl-clang14 libva-wayland2 libxv1 vainfo
0 upgraded, 11 newly installed, 0 to remove and 0 not upgraded.
Need to get 49.2 MB of archives.
After this operation, 254 MB of additional disk space will be used.
Get:1 http://deb.debian.org/debian bookworm/main amd64 libxv1 amd64 2:1.0.11-1.1 [24.8 kB]
Get:2 http://deb.debian.org/debian bookworm/main amd64 intel-gpu-tools amd64 1.27.1-1 [2,339 kB]
Get:3 http://deb.debian.org/debian bookworm/main amd64 libllvm14 amd64 1:14.0.6-12 [21.8 MB]
Get:4 http://deb.debian.org/debian bookworm/main amd64 libllvmspirvlib14 amd64 14.0.0-5 [696 kB]
Get:5 http://deb.debian.org/debian bookworm/main amd64 libigc1 amd64 1.0.12504.6-1+deb12u1 [9,287 kB]
Get:6 http://deb.debian.org/debian bookworm/main amd64 libclang-cpp14 amd64 1:14.0.6-12 [11.1 MB]
Get:7 http://deb.debian.org/debian bookworm/main amd64 libopencl-clang14 amd64 14.0.0-4 [96.1 kB]
Get:8 http://deb.debian.org/debian bookworm/main amd64 libigdfcl1 amd64 1.0.12504.6-1+deb12u1 [86.2 kB]
Get:9 http://deb.debian.org/debian bookworm/main amd64 intel-opencl-icd amd64 22.43.24595.41-1 [3,711 kB]
Get:10 http://deb.debian.org/debian bookworm/main amd64 libva-wayland2 amd64 2.17.0-1 [18.0 kB]
Get:11 http://deb.debian.org/debian bookworm/main amd64 vainfo amd64 2.12.0+ds1-1 [19.2 kB]
Fetched 49.2 MB in 7s (6,664 kB/s)
Selecting previously unselected package libxv1:amd64.
(Reading database ... 48613 files and directories currently installed.)
Preparing to unpack .../00-libxv1_2%3a1.0.11-1.1_amd64.deb ...
Unpacking libxv1:amd64 (2:1.0.11-1.1) ...
Selecting previously unselected package intel-gpu-tools.
Preparing to unpack .../01-intel-gpu-tools_1.27.1-1_amd64.deb ...
Unpacking intel-gpu-tools (1.27.1-1) ...
Selecting previously unselected package libllvm14:amd64.
Preparing to unpack .../02-libllvm14_1%3a14.0.6-12_amd64.deb ...
Unpacking libllvm14:amd64 (1:14.0.6-12) ...
Selecting previously unselected package libllvmspirvlib14:amd64.
Preparing to unpack .../03-libllvmspirvlib14_14.0.0-5_amd64.deb ...
Unpacking libllvmspirvlib14:amd64 (14.0.0-5) ...
Selecting previously unselected package libigc1:amd64.
Preparing to unpack .../04-libigc1_1.0.12504.6-1+deb12u1_amd64.deb ...
Unpacking libigc1:amd64 (1.0.12504.6-1+deb12u1) ...
Selecting previously unselected package libclang-cpp14.
Preparing to unpack .../05-libclang-cpp14_1%3a14.0.6-12_amd64.deb ...
Unpacking libclang-cpp14 (1:14.0.6-12) ...
Selecting previously unselected package libopencl-clang14:amd64.
Preparing to unpack .../06-libopencl-clang14_14.0.0-4_amd64.deb ...
Unpacking libopencl-clang14:amd64 (14.0.0-4) ...
Selecting previously unselected package libigdfcl1:amd64.
Preparing to unpack .../07-libigdfcl1_1.0.12504.6-1+deb12u1_amd64.deb ...
Unpacking libigdfcl1:amd64 (1.0.12504.6-1+deb12u1) ...
Selecting previously unselected package intel-opencl-icd.
Preparing to unpack .../08-intel-opencl-icd_22.43.24595.41-1_amd64.deb ...
Unpacking intel-opencl-icd (22.43.24595.41-1) ...
Selecting previously unselected package libva-wayland2:amd64.
Preparing to unpack .../09-libva-wayland2_2.17.0-1_amd64.deb ...
Unpacking libva-wayland2:amd64 (2.17.0-1) ...
Selecting previously unselected package vainfo.
Preparing to unpack .../10-vainfo_2.12.0+ds1-1_amd64.deb ...
Unpacking vainfo (2.12.0+ds1-1) ...
Setting up libxv1:amd64 (2:1.0.11-1.1) ...
Setting up libva-wayland2:amd64 (2.17.0-1) ...
Setting up libllvm14:amd64 (1:14.0.6-12) ...
Setting up libllvmspirvlib14:amd64 (14.0.0-5) ...
Setting up libigc1:amd64 (1.0.12504.6-1+deb12u1) ...
Setting up intel-gpu-tools (1.27.1-1) ...
Setting up libclang-cpp14 (1:14.0.6-12) ...
Setting up vainfo (2.12.0+ds1-1) ...
Setting up libopencl-clang14:amd64 (14.0.0-4) ...
Setting up libigdfcl1:amd64 (1.0.12504.6-1+deb12u1) ...
Setting up intel-opencl-icd (22.43.24595.41-1) ...
Processing triggers for libc-bin (2.36-9+deb12u13) ...
Processing triggers for man-db (2.11.2-2) ...
chgrp: cannot access '/dev/dri': No such file or directory

[ERROR] in line 25: exit code 0: while executing command chgrp video /dev/dri
/dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context

@remz1337 Please let me know if this is enough or if the entire log is needed.
Thank you for your time

@wlad1164 commented on GitHub (Oct 15, 2025): > [@wlad1164](https://github.com/wlad1164) [@xenoncorp2k](https://github.com/xenoncorp2k) can you run again with verbose output and paste the log back here again? thanks ✔️ Installed Dependencies Reading package lists... Doneeleration Building dependency tree... Done Reading state information... Done va-driver-all is already the newest version (2.17.0-1). va-driver-all set to manually installed. ocl-icd-libopencl1 is already the newest version (2.3.1-1). ocl-icd-libopencl1 set to manually installed. The following additional packages will be installed: libclang-cpp14 libigc1 libigdfcl1 libllvm14 libllvmspirvlib14 libopencl-clang14 libva-wayland2 libxv1 The following NEW packages will be installed: intel-gpu-tools intel-opencl-icd libclang-cpp14 libigc1 libigdfcl1 libllvm14 libllvmspirvlib14 libopencl-clang14 libva-wayland2 libxv1 vainfo 0 upgraded, 11 newly installed, 0 to remove and 0 not upgraded. Need to get 49.2 MB of archives. After this operation, 254 MB of additional disk space will be used. Get:1 http://deb.debian.org/debian bookworm/main amd64 libxv1 amd64 2:1.0.11-1.1 [24.8 kB] Get:2 http://deb.debian.org/debian bookworm/main amd64 intel-gpu-tools amd64 1.27.1-1 [2,339 kB] Get:3 http://deb.debian.org/debian bookworm/main amd64 libllvm14 amd64 1:14.0.6-12 [21.8 MB] Get:4 http://deb.debian.org/debian bookworm/main amd64 libllvmspirvlib14 amd64 14.0.0-5 [696 kB] Get:5 http://deb.debian.org/debian bookworm/main amd64 libigc1 amd64 1.0.12504.6-1+deb12u1 [9,287 kB] Get:6 http://deb.debian.org/debian bookworm/main amd64 libclang-cpp14 amd64 1:14.0.6-12 [11.1 MB] Get:7 http://deb.debian.org/debian bookworm/main amd64 libopencl-clang14 amd64 14.0.0-4 [96.1 kB] Get:8 http://deb.debian.org/debian bookworm/main amd64 libigdfcl1 amd64 1.0.12504.6-1+deb12u1 [86.2 kB] Get:9 http://deb.debian.org/debian bookworm/main amd64 intel-opencl-icd amd64 22.43.24595.41-1 [3,711 kB] Get:10 http://deb.debian.org/debian bookworm/main amd64 libva-wayland2 amd64 2.17.0-1 [18.0 kB] Get:11 http://deb.debian.org/debian bookworm/main amd64 vainfo amd64 2.12.0+ds1-1 [19.2 kB] Fetched 49.2 MB in 7s (6,664 kB/s) Selecting previously unselected package libxv1:amd64. (Reading database ... 48613 files and directories currently installed.) Preparing to unpack .../00-libxv1_2%3a1.0.11-1.1_amd64.deb ... Unpacking libxv1:amd64 (2:1.0.11-1.1) ... Selecting previously unselected package intel-gpu-tools. Preparing to unpack .../01-intel-gpu-tools_1.27.1-1_amd64.deb ... Unpacking intel-gpu-tools (1.27.1-1) ... Selecting previously unselected package libllvm14:amd64. Preparing to unpack .../02-libllvm14_1%3a14.0.6-12_amd64.deb ... Unpacking libllvm14:amd64 (1:14.0.6-12) ... Selecting previously unselected package libllvmspirvlib14:amd64. Preparing to unpack .../03-libllvmspirvlib14_14.0.0-5_amd64.deb ... Unpacking libllvmspirvlib14:amd64 (14.0.0-5) ... Selecting previously unselected package libigc1:amd64. Preparing to unpack .../04-libigc1_1.0.12504.6-1+deb12u1_amd64.deb ... Unpacking libigc1:amd64 (1.0.12504.6-1+deb12u1) ... Selecting previously unselected package libclang-cpp14. Preparing to unpack .../05-libclang-cpp14_1%3a14.0.6-12_amd64.deb ... Unpacking libclang-cpp14 (1:14.0.6-12) ... Selecting previously unselected package libopencl-clang14:amd64. Preparing to unpack .../06-libopencl-clang14_14.0.0-4_amd64.deb ... Unpacking libopencl-clang14:amd64 (14.0.0-4) ... Selecting previously unselected package libigdfcl1:amd64. Preparing to unpack .../07-libigdfcl1_1.0.12504.6-1+deb12u1_amd64.deb ... Unpacking libigdfcl1:amd64 (1.0.12504.6-1+deb12u1) ... Selecting previously unselected package intel-opencl-icd. Preparing to unpack .../08-intel-opencl-icd_22.43.24595.41-1_amd64.deb ... Unpacking intel-opencl-icd (22.43.24595.41-1) ... Selecting previously unselected package libva-wayland2:amd64. Preparing to unpack .../09-libva-wayland2_2.17.0-1_amd64.deb ... Unpacking libva-wayland2:amd64 (2.17.0-1) ... Selecting previously unselected package vainfo. Preparing to unpack .../10-vainfo_2.12.0+ds1-1_amd64.deb ... Unpacking vainfo (2.12.0+ds1-1) ... Setting up libxv1:amd64 (2:1.0.11-1.1) ... Setting up libva-wayland2:amd64 (2.17.0-1) ... Setting up libllvm14:amd64 (1:14.0.6-12) ... Setting up libllvmspirvlib14:amd64 (14.0.0-5) ... Setting up libigc1:amd64 (1.0.12504.6-1+deb12u1) ... Setting up intel-gpu-tools (1.27.1-1) ... Setting up libclang-cpp14 (1:14.0.6-12) ... Setting up vainfo (2.12.0+ds1-1) ... Setting up libopencl-clang14:amd64 (14.0.0-4) ... Setting up libigdfcl1:amd64 (1.0.12504.6-1+deb12u1) ... Setting up intel-opencl-icd (22.43.24595.41-1) ... Processing triggers for libc-bin (2.36-9+deb12u13) ... Processing triggers for man-db (2.11.2-2) ... chgrp: cannot access '/dev/dri': No such file or directory [ERROR] in line 25: exit code 0: while executing command chgrp video /dev/dri /dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context @remz1337 Please let me know if this is enough or if the entire log is needed. Thank you for your time

kerem commented

Author

Owner

@xenoncorp2k commented on GitHub (Oct 15, 2025):

Today I test with: bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"

And it works. No Building HailoRT error.

@xenoncorp2k commented on GitHub (Oct 15, 2025): Today I test with: bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)" And it works. No Building HailoRT error.

kerem commented

Author

Owner

@crhy commented on GitHub (Oct 16, 2025):

This is what I got with the link xeno just posted:

➤ Automatically mount all available VAAPI devices? [Y/n]: y
✔️ Started LXC Container
💡 No network in LXC yet (try 1/10) – waiting...
✔️ Network in LXC is reachable (ping)
✔️ Customized LXC Container
Customizing LXC creation
✔️ Rebooted LXC
✔️ Set up Container OS
✔️ Network Connected: 192.168.88.22
✔️ IPv4 Internet Connected
✖️ IPv6 Internet Not Connected
✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ )
api.github.com:(✔️ )
✔️ Updated Container OS
✔️ Set up Container
✔️ Installed APT proxy client
✔️ Installed sudo
/dev/fd/62: line 134: pop_var_context: head of shell_variables not a
function context
@.***:~#

On Wed, Oct 15, 2025 at 10:10 AM xenoncorp2k @.***>
wrote:

xenoncorp2k left a comment (community-scripts/ProxmoxVE#6915)
https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3408076900

Today I test with: bash -c "$(wget -qLO -
https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"

And it works. No Building HailoRT error.

—
Reply to this email directly, view it on GitHub
https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3408076900,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AFBTXI76HCD6KBEPE7WLB233X2SZXAVCNFSM6AAAAACEC6ZHIWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTIMBYGA3TMOJQGA
.
You are receiving this because you were mentioned.Message ID:
@.***>

--

Cowboy Crooner, Violinist, Composer, Linux Guy, World's Best Dad, AirShip
Designer

rhymusic.com
OpenAirShips.com
ThorntonTechs.com
CoinKitty.org
SuperMox.me
https://leperkhanz.weebly.com/ http://leperkhanz.weebly.com/

@crhy commented on GitHub (Oct 16, 2025): This is what I got with the link xeno just posted: ➤ Automatically mount all available VAAPI devices? [Y/n]: y ✔️ Started LXC Container 💡 No network in LXC yet (try 1/10) – waiting... ✔️ Network in LXC is reachable (ping) ✔️ Customized LXC Container Customizing LXC creation ✔️ Rebooted LXC ✔️ Set up Container OS ✔️ Network Connected: 192.168.88.22 ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) ✔️ Updated Container OS ✔️ Set up Container ✔️ Installed APT proxy client ✔️ Installed sudo /dev/fd/62: line 134: pop_var_context: head of shell_variables not a function context ***@***.***:~# On Wed, Oct 15, 2025 at 10:10 AM xenoncorp2k ***@***.***> wrote: > *xenoncorp2k* left a comment (community-scripts/ProxmoxVE#6915) > <https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3408076900> > > Today I test with: bash -c "$(wget -qLO - > https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)" > > And it works. No Building HailoRT error. > > — > Reply to this email directly, view it on GitHub > <https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3408076900>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AFBTXI76HCD6KBEPE7WLB233X2SZXAVCNFSM6AAAAACEC6ZHIWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTIMBYGA3TMOJQGA> > . > You are receiving this because you were mentioned.Message ID: > ***@***.***> > -- - Cowboy Crooner, Violinist, Composer, Linux Guy, World's Best Dad, AirShip Designer rhymusic.com OpenAirShips.com ThorntonTechs.com CoinKitty.org SuperMox.me https://leperkhanz.weebly.com/ <http://leperkhanz.weebly.com/>

kerem commented

Author

Owner

@simplemice commented on GitHub (Oct 16, 2025):

✔️ Installed APT proxy client
✔️ Installed sudo
NVIDIA drivers not found!
⚠ User exited script

@simplemice commented on GitHub (Oct 16, 2025): ✔️ Installed APT proxy client ✔️ Installed sudo NVIDIA drivers not found! ⚠ User exited script

kerem commented

Author

Owner

@remz1337 commented on GitHub (Oct 16, 2025):

for iGPU : --> advanced settings: make it priviledged and disable Nvidia GPU passthrough
for Nvidia: ensure drivers are installed on the host (test with nvidia-smi on the host), then run the script with default settings

@remz1337 commented on GitHub (Oct 16, 2025): for iGPU : --> advanced settings: make it priviledged and disable Nvidia GPU passthrough for Nvidia: ensure drivers are installed on the host (test with `nvidia-smi` on the host), then run the script with default settings

kerem commented

Author

Owner

@remz1337 commented on GitHub (Oct 16, 2025):

If you are still facing issues, try a simple CPU frigate installation (Unpriviledged, no VAAPI, no Nvidia)

@remz1337 commented on GitHub (Oct 16, 2025): If you are still facing issues, try a simple CPU frigate installation (Unpriviledged, no VAAPI, no Nvidia)

kerem commented

Author

Owner

@xenoncorp2k commented on GitHub (Oct 16, 2025):

All working except HA integration. Solution rename config.yml to config.yaml and works too.

Install with priviledged, yes VAAPI, no Nvidia. iGPU hardware acceleration OK. 8-9ms inference speed with i7-7700.

Thanks for all.

@xenoncorp2k commented on GitHub (Oct 16, 2025): All working except HA integration. Solution rename config.yml to config.yaml and works too. Install with priviledged, yes VAAPI, no Nvidia. iGPU hardware acceleration OK. 8-9ms inference speed with i7-7700. Thanks for all.

kerem commented

Author

Owner

@ask2ward commented on GitHub (Oct 16, 2025):

Hello!
I test with : bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)" with remz1337 comment : If Unpriviledged, no VAAPI, no Nvidia.
And it's not working
Here's the log:
[ INFO ] MO command line tool is considered as the legacy conversion API as of OpenVINO 2023.2 release.
In 2025.0 MO command line tool and openvino.tools.mo.convert_model() will be removed. Please use OpenVINO Model Converter (OVC) or openvino.convert_model(). OVC represents a lightweight alternative of MO and provides simplified model conversion API.
Find more information about transition from MO to OVC at https://docs.openvino.ai/2023.2/openvino_docs_OV_Converter_UG_prepare_model_convert_model_MO_OVC_transition.html
0: line 189: 31227 Illegal instruction $STD python3 /opt/frigate/docker/main/build_ov_model.py

[ERROR] in line 189: exit code 0: while executing command $STD python3 /opt/frigate/docker/main/build_ov_model.py
/dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context
I have proxmox v 9.0.11

Thanks

@ask2ward commented on GitHub (Oct 16, 2025): Hello! I test with : bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)" with remz1337 comment : If Unpriviledged, no VAAPI, no Nvidia. And it's not working Here's the log: [ INFO ] MO command line tool is considered as the legacy conversion API as of OpenVINO 2023.2 release. In 2025.0 MO command line tool and openvino.tools.mo.convert_model() will be removed. Please use OpenVINO Model Converter (OVC) or openvino.convert_model(). OVC represents a lightweight alternative of MO and provides simplified model conversion API. Find more information about transition from MO to OVC at https://docs.openvino.ai/2023.2/openvino_docs_OV_Converter_UG_prepare_model_convert_model_MO_OVC_transition.html 0: line 189: 31227 Illegal instruction $STD python3 /opt/frigate/docker/main/build_ov_model.py [ERROR] in line 189: exit code 0: while executing command $STD python3 /opt/frigate/docker/main/build_ov_model.py /dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context I have proxmox v 9.0.11 Thanks

kerem commented

Author

Owner

@AsmoOne commented on GitHub (Oct 16, 2025):

@remz1337 Hi! I am set privilegied container, vaapi yes, and passthrou intel GPU. Sp latest log is:

tar: Ignoring unknown extended header keyword 'LIBARCHIVE.xattr.com.apple.quarantine'
tar: Ignoring unknown extended header keyword 'LIBARCHIVE.xattr.com.apple.quarantine'

rm -rf ffmpeg.tar.xz
mkdir -p /usr/lib/ffmpeg/7.0
wget -qO ffmpeg.tar.xz https://github.com/NickM-27/FFmpeg-Builds/releases/download/autobuild-2024-09-19-12-51/ffmpeg-n7.0.2-18-g3e6cec1286-linux64-gpl-7.0.tar.xz
tar -xf ffmpeg.tar.xz -C /usr/lib/ffmpeg/7.0 --strip-components 1 amd64/bin/ffmpeg amd64/bin/ffprobe
rm -rf ffmpeg.tar.xz
sed -i -E '/^Components: main$/s/main/main contrib non-free non-free-firmware/' /etc/apt/sources.list.d/debian.sources
sed: can't read /etc/apt/sources.list.d/debian.sources: No such file or directory
apt-get -qq install --no-install-recommends --no-install-suggests -y intel-gpu-tools onevpl-tools libva-drm2 mesa-va-drivers radeontop
Selecting previously unselected package libvpl2.
(Reading database ... 66945 files and directories currently installed.)
Preparing to unpack .../libvpl2_2023.1.1-1_amd64.deb ...
Unpacking libvpl2 (2023.1.1-1) ...
Selecting previously unselected package onevpl-tools.
Preparing to unpack .../onevpl-tools_2023.1.1-1_amd64.deb ...
Unpacking onevpl-tools (2023.1.1-1) ...
Selecting previously unselected package radeontop.
Preparing to unpack .../radeontop_1.4-2_amd64.deb ...
Unpacking radeontop (1.4-2) ...
Setting up libvpl2 (2023.1.1-1) ...
Setting up onevpl-tools (2023.1.1-1) ...
Setting up radeontop (1.4-2) ...
Processing triggers for man-db (2.11.2-2) ...
Processing triggers for libc-bin (2.36-9+deb12u13) ...
apt-get install -y dpkg
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
dpkg is already the newest version (1.21.22).
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
wget -qO - https://repositories.intel.com/gpu/intel-graphics.key
gpg --yes --dearmor --output /usr/share/keyrings/intel-graphics.gpg
gpg: no valid OpenPGP data found.

[ERROR] in line 173: exit code 0: while executing command $STD bash /opt/frigate/docker/main/install_deps.sh
/dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context

@AsmoOne commented on GitHub (Oct 16, 2025): @remz1337 Hi! I am set privilegied container, vaapi yes, and passthrou intel GPU. Sp latest log is: tar: Ignoring unknown extended header keyword 'LIBARCHIVE.xattr.com.apple.quarantine' tar: Ignoring unknown extended header keyword 'LIBARCHIVE.xattr.com.apple.quarantine' + rm -rf ffmpeg.tar.xz + mkdir -p /usr/lib/ffmpeg/7.0 + wget -qO ffmpeg.tar.xz https://github.com/NickM-27/FFmpeg-Builds/releases/download/autobuild-2024-09-19-12-51/ffmpeg-n7.0.2-18-g3e6cec1286-linux64-gpl-7.0.tar.xz + tar -xf ffmpeg.tar.xz -C /usr/lib/ffmpeg/7.0 --strip-components 1 amd64/bin/ffmpeg amd64/bin/ffprobe + rm -rf ffmpeg.tar.xz + [[ amd64 == \a\r\m\6\4 ]] + [[ amd64 == \a\m\d\6\4 ]] + sed -i -E '/^Components: main$/s/main/main contrib non-free non-free-firmware/' /etc/apt/sources.list.d/debian.sources sed: can't read /etc/apt/sources.list.d/debian.sources: No such file or directory + apt-get -qq install --no-install-recommends --no-install-suggests -y intel-gpu-tools onevpl-tools libva-drm2 mesa-va-drivers radeontop Selecting previously unselected package libvpl2. (Reading database ... 66945 files and directories currently installed.) Preparing to unpack .../libvpl2_2023.1.1-1_amd64.deb ... Unpacking libvpl2 (2023.1.1-1) ... Selecting previously unselected package onevpl-tools. Preparing to unpack .../onevpl-tools_2023.1.1-1_amd64.deb ... Unpacking onevpl-tools (2023.1.1-1) ... Selecting previously unselected package radeontop. Preparing to unpack .../radeontop_1.4-2_amd64.deb ... Unpacking radeontop (1.4-2) ... Setting up libvpl2 (2023.1.1-1) ... Setting up onevpl-tools (2023.1.1-1) ... Setting up radeontop (1.4-2) ... Processing triggers for man-db (2.11.2-2) ... Processing triggers for libc-bin (2.36-9+deb12u13) ... + apt-get install -y dpkg Reading package lists... Done Building dependency tree... Done Reading state information... Done dpkg is already the newest version (1.21.22). 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. + wget -qO - https://repositories.intel.com/gpu/intel-graphics.key + gpg --yes --dearmor --output /usr/share/keyrings/intel-graphics.gpg gpg: no valid OpenPGP data found. [ERROR] in line 173: exit code 0: while executing command $STD bash /opt/frigate/docker/main/install_deps.sh /dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context

kerem commented

Author

Owner

@Lennart95 commented on GitHub (Oct 17, 2025):

Not 100% sure if it is related to the installation using this new script or mentioned before, but restarting through the frigate UI isn't working for me anymore. The spinner continues to spin as if it is still starting up. None of the tabs can be accessed, reboot of the LXC container is needed to come back up.

@Lennart95 commented on GitHub (Oct 17, 2025): Not 100% sure if it is related to the installation using this new script or mentioned before, but restarting through the frigate UI isn't working for me anymore. The spinner continues to spin as if it is still starting up. None of the tabs can be accessed, reboot of the LXC container is needed to come back up.

kerem commented

Author

Owner

@Zixim commented on GitHub (Oct 17, 2025):

@Lennart95 I ran into that when I added face recognition to my config.

@Zixim commented on GitHub (Oct 17, 2025): @Lennart95 I ran into that when I added face recognition to my config.

kerem commented

Author

Owner

@Lennart95 commented on GitHub (Oct 18, 2025):

@Lennart95 I ran into that when I added face recognition to my config.

Commenting out the config doesn't make any difference in my case.

@Lennart95 commented on GitHub (Oct 18, 2025): > [@Lennart95](https://github.com/Lennart95) I ran into that when I added face recognition to my config. Commenting out the config doesn't make any difference in my case.

kerem commented

Author

Owner

@sorinbidian-cell commented on GitHub (Oct 23, 2025):

Nevermind.
I tested bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)" on proxmox 8.4.14
Working with Coral TPU M.2
Frigate is not restarting. Stuck in peewee.sqliteq INFO : writer received shutdown request, exiting
A can restart using systemctl restart frigate.service.
Also my Hikvision cameras give a lot of errors with ffmepg 7.0 and I switch to 5.0

@sorinbidian-cell commented on GitHub (Oct 23, 2025): Nevermind. I tested bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)" on proxmox 8.4.14 Working with Coral TPU M.2 Frigate is not restarting. Stuck in peewee.sqliteq INFO : writer received shutdown request, exiting A can restart using systemctl restart frigate.service. Also my Hikvision cameras give a lot of errors with ffmepg 7.0 and I switch to 5.0

kerem commented

Author

Owner

@timhbt commented on GitHub (Oct 30, 2025):

bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVED/main/ct/frigate.sh)"

privileged, no nvidia passthrough, vaapi

Failed to build nvidia-pyindex
error: failed-wheel-build-for-install

× Failed to build installable wheels for some pyproject.toml based projects
╰─> nvidia-pyindex

[ERROR] in line 126: exit code 0: while executing command $STD pip3 install -r /opt/frigate/docker/main/requirements.txt
/dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context

Edit:
Should be fixed in a future frigate release:
https://github.com/blakeblackshear/frigate/pull/20666

@timhbt commented on GitHub (Oct 30, 2025): `bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVED/main/ct/frigate.sh)"` **privileged, no nvidia passthrough, vaapi** ``` Failed to build nvidia-pyindex error: failed-wheel-build-for-install × Failed to build installable wheels for some pyproject.toml based projects ╰─> nvidia-pyindex [ERROR] in line 126: exit code 0: while executing command $STD pip3 install -r /opt/frigate/docker/main/requirements.txt /dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context ``` **Edit:** Should be fixed in a future frigate release: https://github.com/blakeblackshear/frigate/pull/20666

kerem commented

Author

Owner

@ShaMAD commented on GitHub (Nov 1, 2025):

Hi, I'm having trouble trying to use Google Coral TPU in an unprivileged LXC Frigate won't start.
At the same time, access for unprivileged LXC is configured on the host https://github.com/Bodengriller/frigate_debian_scripts/blob/Proxmox-Edition/install_coral_tpu.sh

root@pve-s1:/dev# ll | grep apex
crwxrwxrwx  1 root apex    120,     0 Nov  1 12:04 apex_0
crwxrwxrwx  1 root apex    120,     1 Nov  1 12:04 apex_1

root@pve-s1:/dev#  cd && ./install_coral_tpu.sh --status
Coral TPU PCIe hardware detected:
03:00.0 System peripheral [0880]: Global Unichip Corp. Coral Edge TPU [1ac1:089a]
04:00.0 System peripheral [0880]: Global Unichip Corp. Coral Edge TPU [1ac1:089a]

Apex devices installed:
/dev/apex_0
/dev/apex_1

DKMS Gasket module: gasket/1.0, 6.12.43+deb13-amd64, x86_64: installed
gasket/1.0, 6.12.48+deb13-amd64, x86_64: installed
gasket/1.0, 6.14.11-1-pve, x86_64: installed
gasket/1.0, 6.14.11-2-pve, x86_64: installed
gasket/1.0, 6.14.11-4-pve, x86_64: installed
gasket/1.0, 6.14.8-2-pve, x86_64: installed
Apex module version: 1.2
libedgetpu1-std version: 16.0

Non-root access:
Udev rule for non-root access: present.
Apex group for non-root access: present.
Members: root

Unprivileged LXC access:
Udev rule for unprivileged LXC access: present.

root@pve-s1:/etc/pve/lxc# cat 210.conf
arch: amd64
cores: 4
dev0: /dev/dri/card0,gid=44
dev1: /dev/dri/renderD128,gid=104
dev2: /dev/apex_0,gid=1000
dev3: /dev/apex_1,gid=1000
features: nesting=1
hostname: lg-frigate
memory: 4096
mp0: /mnt/Media/NVRFrigate,mp=/media/frigate
net0: name=eth0,bridge=vmbr0,firewall=1,hwaddr=BC:24:11:CD:58:9C,ip=dhcp,ip6=auto,type=veth
onboot: 1
ostype: debian
rootfs: Data:vm-210-disk-0,size=50G
startup: order=5,up=10
swap: 4096
tags: 5;nvr
unprivileged: 1

And configure detectors

root@lg-frigate:/config# cat config.yml | grep -i -A 15 "detectors"
detectors:
  coral1:
    type: edgetpu
    device: pci:0
  coral2:
    type: edgetpu
    device: pci:1
  detector01:
    type: openvino
model:
  width: 300
  height: 300
  input_tensor: nhwc
  input_pixel_format: bgr
  path: /openvino-model/ssdlite_mobilenet_v2.xml
  labelmap_path: /openvino-model/coco_91cl_bkgr.txt

root@lg-frigate:/config# ls /
audio-labelmap.txt  config                etc       labelmap.txt  lost+found  models          proc    run   sys  v0.1.3.tar.gz
bin                 dev                   home      lib           media       openvino-model  root    sbin  tmp  var
boot                edgetpu_model.tflite  labelmap  lib64         mnt         opt             rootfs  srv   usr  wheels

But the frigate won't start.

root@lg-frigate:~# cat /dev/shm/logs/frigate/current | grep tpu
2025-11-01 12:09:48.159783  [2025-11-01 12:09:48] frigate.detectors.plugins.edgetpu_tfl INFO    : Attempting to load TPU as pci:0
2025-11-01 12:09:48.172407  [2025-11-01 12:09:48] frigate.detectors.plugins.edgetpu_tfl INFO    : Attempting to load TPU as pci:1
2025-11-01 12:09:48.173729  [2025-11-01 12:09:48] frigate.detectors.plugins.edgetpu_tfl INFO    : TPU found
2025-11-01 12:09:48.175064  [2025-11-01 12:09:48] frigate.detectors.plugins.edgetpu_tfl INFO    : TPU found
2025-11-01 12:09:48.186004  [2025-11-01 12:09:48] frigate.detectors.plugins.edgetpu_tfl ERROR   : Incorrect model used with EdgeTPU. Only .tflite models can be used with a Coral EdgeTPU.
2025-11-01 12:09:48.189082    File "/opt/frigate/frigate/detectors/plugins/edgetpu_tfl.py", line 44, in __init__
2025-11-01 12:09:48.189534    File "/opt/frigate/frigate/detectors/plugins/edgetpu_tfl.py", line 44, in __init__
2025-11-01 12:09:48.189661  [2025-11-01 12:09:48] frigate.detectors.plugins.edgetpu_tfl ERROR   : Incorrect model used with EdgeTPU. Only .tflite models can be used with a Coral EdgeTPU.
2025-11-01 12:09:48.189689  [2025-11-01 12:09:48] frigate.app                    INFO    : Output process started: 641
2025-11-01 12:10:22.854811  [2025-11-01 12:10:22] frigate.output.output          INFO    : exiting output process...

I tried installing privileged LXC but I get an error

  🧩  Using Advanced Install on node pve-s1
  💡  PVE Version 9.0.11 (Kernel: 6.14.11-2-pve)
  💡  PVE Version 9.0.11 (Kernel: 6.14.11-2-pve)
  🖥️  Operating System: debian
  🌟  Version: 13
  📦  Container Type: Privileged
  🔐  Root Password: ********
  🆔  Container ID: 253
  🏠  Hostname: lg-frigate
  💾  Disk Size: 32 GB
  🧠  CPU Cores: 4
  🛠️  RAM Size: 4096 MiB
  🌉  Bridge: vmbr0
  📡  IPv4: DHCP
  📡  IPv6: SLAAC/AUTO
  📡  APT-Cacher IP Address: Default
  ⚙️  Interface MTU Size: Default
  🔍  DNS Search Domain: Host
  📡  DNS Server IP Address: Host
  🏷️  Vlan: Default
  📡  Tags: community-script;nvr
  🔑  Root SSH Access: yes
  🗂️  Enable FUSE Support: no
  🔍  Verbose Mode: yes
  🚀  Creating a Frigate LXC using the above advanced settings
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
/dev/fd/63: line 1241: _is_whitelisted_key: command not found
  ✔️   Storage local (Free: 64.8GB  Used: 22.3GB) [Template]
  ✔️   Storage Data (Free: 203.6GB  Used: 82.6GB) [Container]
  ⏳   Searching for template 'debian-13'  - debian-13-standard_13.1-2_amd64.tar.zst
  ✔️   Template debian-13-standard_13.1-2_amd64.tar.zst [online]
  ✔️   LXC Container 253 was successfully created.
  ✔️   USB passthrough configured
  ✔️   INTEL GPU passthrough configured (2 devices)
  ✔️   Started LXC Container
   💡   Network reachable but gateway check failed
  ✔️   Device GIDs updated successfully
  ✔️   Customized LXC Container
  ✔️   Installed SSH keys into CT 253
  ✔️   Set up Container OS
  ✔️   Network Connected: 172.17.15.242
  ✔️   IPv4 Internet Connected
   ✖️   IPv6 Internet Not Connected
  ✔️   Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) git.community-scripts.org:(✔️ )
Hit:1 http://deb.debian.org/debian trixie InRelease
Hit:2 http://security.debian.org trixie-security InRelease
Hit:3 http://deb.debian.org/debian trixie-updates InRelease
Reading package lists... Done
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
Calculating upgrade... Done
The following packages will be upgraded:
  bind9-dnsutils bind9-host bind9-libs
3 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
Need to get 1,459 kB of archives.
After this operation, 24.6 kB of additional disk space will be used.
Get:1 http://security.debian.org trixie-security/main amd64 bind9-host amd64 1:9.20.15-1~deb13u1 [55.4 kB]
Get:2 http://security.debian.org trixie-security/main amd64 bind9-dnsutils amd64 1:9.20.15-1~deb13u1 [165 kB]
Get:3 http://security.debian.org trixie-security/main amd64 bind9-libs amd64 1:9.20.15-1~deb13u1 [1,238 kB]
Fetched 1,459 kB in 1s (1,168 kB/s)
apt-listchanges: Reading changelogs...
(Reading database ... 20272 files and directories currently installed.)
Preparing to unpack .../bind9-host_1%3a9.20.15-1~deb13u1_amd64.deb ...
Unpacking bind9-host (1:9.20.15-1~deb13u1) over (1:9.20.11-4) ...
Preparing to unpack .../bind9-dnsutils_1%3a9.20.15-1~deb13u1_amd64.deb ...
Unpacking bind9-dnsutils (1:9.20.15-1~deb13u1) over (1:9.20.11-4) ...
Preparing to unpack .../bind9-libs_1%3a9.20.15-1~deb13u1_amd64.deb ...
Unpacking bind9-libs:amd64 (1:9.20.15-1~deb13u1) over (1:9.20.11-4) ...
Setting up bind9-libs:amd64 (1:9.20.15-1~deb13u1) ...
Setting up bind9-host (1:9.20.15-1~deb13u1) ...
Setting up bind9-dnsutils (1:9.20.15-1~deb13u1) ...
Processing triggers for man-db (2.13.1-1) ...
Processing triggers for libc-bin (2.41-12) ...
  ✔️   Updated Container OS
Reading package lists... Dones (Patience)
Building dependency tree... Done
Reading state information... Done
Note, selecting 'tcl' instead of 'tclsh'
Package libatlas-base-dev is not available, but is referred to by another package.
This may mean that the package is missing, has been obsoleted, or
is only available from another source

Package python3-distutils is not available, but is referred to by another package.
This may mean that the package is missing, has been obsoleted, or
is only available from another source

E: Package 'python3-distutils' has no installation candidate
E: Unable to locate package python3.11
E: Couldn't find any package by glob 'python3.11'
E: Couldn't find any package by regex 'python3.11'
E: Unable to locate package python3.11-dev
E: Couldn't find any package by glob 'python3.11-dev'
E: Couldn't find any package by regex 'python3.11-dev'
E: Package 'libatlas-base-dev' has no installation candidate

[ERROR] in line 18: exit code 18 (Unknown error): while executing command  apt-get install -y {jq,wget,xz-utils,python3,python3-dev,python3-distutils,gcc,pkg-config,libhdf5-dev,unzip,build-essential,automake,libtool,ccache,libusb-1.0-0-dev,apt-transport-https,python3.11,python3.11-dev,cmake,git,libgtk-3-dev,libavcodec-dev,libavformat-dev,libswscale-dev,libv4l-dev,libxvidcore-dev,libx264-dev,libjpeg-dev,libpng-dev,libtiff-dev,gfortran,openexr,libatlas-base-dev,libssl-dev,libtbbmalloc2,libtbb-dev,libdc1394-dev,libopenexr-dev,libgstreamer-plugins-base1.0-dev,libgstreamer1.0-dev,tclsh,libopenblas-dev,liblapack-dev,make,moreutils}

@ShaMAD commented on GitHub (Nov 1, 2025): Hi, I'm having trouble trying to use Google Coral TPU in an unprivileged LXC Frigate won't start. At the same time, access for unprivileged LXC is configured on the host https://github.com/Bodengriller/frigate_debian_scripts/blob/Proxmox-Edition/install_coral_tpu.sh ``` root@pve-s1:/dev# ll | grep apex crwxrwxrwx 1 root apex 120, 0 Nov 1 12:04 apex_0 crwxrwxrwx 1 root apex 120, 1 Nov 1 12:04 apex_1 root@pve-s1:/dev# cd && ./install_coral_tpu.sh --status Coral TPU PCIe hardware detected: 03:00.0 System peripheral [0880]: Global Unichip Corp. Coral Edge TPU [1ac1:089a] 04:00.0 System peripheral [0880]: Global Unichip Corp. Coral Edge TPU [1ac1:089a] Apex devices installed: /dev/apex_0 /dev/apex_1 DKMS Gasket module: gasket/1.0, 6.12.43+deb13-amd64, x86_64: installed gasket/1.0, 6.12.48+deb13-amd64, x86_64: installed gasket/1.0, 6.14.11-1-pve, x86_64: installed gasket/1.0, 6.14.11-2-pve, x86_64: installed gasket/1.0, 6.14.11-4-pve, x86_64: installed gasket/1.0, 6.14.8-2-pve, x86_64: installed Apex module version: 1.2 libedgetpu1-std version: 16.0 Non-root access: Udev rule for non-root access: present. Apex group for non-root access: present. Members: root Unprivileged LXC access: Udev rule for unprivileged LXC access: present. ``` ``` root@pve-s1:/etc/pve/lxc# cat 210.conf arch: amd64 cores: 4 dev0: /dev/dri/card0,gid=44 dev1: /dev/dri/renderD128,gid=104 dev2: /dev/apex_0,gid=1000 dev3: /dev/apex_1,gid=1000 features: nesting=1 hostname: lg-frigate memory: 4096 mp0: /mnt/Media/NVRFrigate,mp=/media/frigate net0: name=eth0,bridge=vmbr0,firewall=1,hwaddr=BC:24:11:CD:58:9C,ip=dhcp,ip6=auto,type=veth onboot: 1 ostype: debian rootfs: Data:vm-210-disk-0,size=50G startup: order=5,up=10 swap: 4096 tags: 5;nvr unprivileged: 1 ``` And configure detectors ``` root@lg-frigate:/config# cat config.yml | grep -i -A 15 "detectors" detectors: coral1: type: edgetpu device: pci:0 coral2: type: edgetpu device: pci:1 detector01: type: openvino model: width: 300 height: 300 input_tensor: nhwc input_pixel_format: bgr path: /openvino-model/ssdlite_mobilenet_v2.xml labelmap_path: /openvino-model/coco_91cl_bkgr.txt root@lg-frigate:/config# ls / audio-labelmap.txt config etc labelmap.txt lost+found models proc run sys v0.1.3.tar.gz bin dev home lib media openvino-model root sbin tmp var boot edgetpu_model.tflite labelmap lib64 mnt opt rootfs srv usr wheels ``` But the frigate won't start. ``` root@lg-frigate:~# cat /dev/shm/logs/frigate/current | grep tpu 2025-11-01 12:09:48.159783 [2025-11-01 12:09:48] frigate.detectors.plugins.edgetpu_tfl INFO : Attempting to load TPU as pci:0 2025-11-01 12:09:48.172407 [2025-11-01 12:09:48] frigate.detectors.plugins.edgetpu_tfl INFO : Attempting to load TPU as pci:1 2025-11-01 12:09:48.173729 [2025-11-01 12:09:48] frigate.detectors.plugins.edgetpu_tfl INFO : TPU found 2025-11-01 12:09:48.175064 [2025-11-01 12:09:48] frigate.detectors.plugins.edgetpu_tfl INFO : TPU found 2025-11-01 12:09:48.186004 [2025-11-01 12:09:48] frigate.detectors.plugins.edgetpu_tfl ERROR : Incorrect model used with EdgeTPU. Only .tflite models can be used with a Coral EdgeTPU. 2025-11-01 12:09:48.189082 File "/opt/frigate/frigate/detectors/plugins/edgetpu_tfl.py", line 44, in __init__ 2025-11-01 12:09:48.189534 File "/opt/frigate/frigate/detectors/plugins/edgetpu_tfl.py", line 44, in __init__ 2025-11-01 12:09:48.189661 [2025-11-01 12:09:48] frigate.detectors.plugins.edgetpu_tfl ERROR : Incorrect model used with EdgeTPU. Only .tflite models can be used with a Coral EdgeTPU. 2025-11-01 12:09:48.189689 [2025-11-01 12:09:48] frigate.app INFO : Output process started: 641 2025-11-01 12:10:22.854811 [2025-11-01 12:10:22] frigate.output.output INFO : exiting output process... ``` I tried installing privileged LXC but I get an error ``` 🧩 Using Advanced Install on node pve-s1 💡 PVE Version 9.0.11 (Kernel: 6.14.11-2-pve) 💡 PVE Version 9.0.11 (Kernel: 6.14.11-2-pve) 🖥️ Operating System: debian 🌟 Version: 13 📦 Container Type: Privileged 🔐 Root Password: ******** 🆔 Container ID: 253 🏠 Hostname: lg-frigate 💾 Disk Size: 32 GB 🧠 CPU Cores: 4 🛠️ RAM Size: 4096 MiB 🌉 Bridge: vmbr0 📡 IPv4: DHCP 📡 IPv6: SLAAC/AUTO 📡 APT-Cacher IP Address: Default ⚙️ Interface MTU Size: Default 🔍 DNS Search Domain: Host 📡 DNS Server IP Address: Host 🏷️ Vlan: Default 📡 Tags: community-script;nvr 🔑 Root SSH Access: yes 🗂️ Enable FUSE Support: no 🔍 Verbose Mode: yes 🚀 Creating a Frigate LXC using the above advanced settings /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found /dev/fd/63: line 1241: _is_whitelisted_key: command not found ✔️ Storage local (Free: 64.8GB Used: 22.3GB) [Template] ✔️ Storage Data (Free: 203.6GB Used: 82.6GB) [Container] ⏳ Searching for template 'debian-13' - debian-13-standard_13.1-2_amd64.tar.zst ✔️ Template debian-13-standard_13.1-2_amd64.tar.zst [online] ✔️ LXC Container 253 was successfully created. ✔️ USB passthrough configured ✔️ INTEL GPU passthrough configured (2 devices) ✔️ Started LXC Container 💡 Network reachable but gateway check failed ✔️ Device GIDs updated successfully ✔️ Customized LXC Container ✔️ Installed SSH keys into CT 253 ✔️ Set up Container OS ✔️ Network Connected: 172.17.15.242 ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) git.community-scripts.org:(✔️ ) Hit:1 http://deb.debian.org/debian trixie InRelease Hit:2 http://security.debian.org trixie-security InRelease Hit:3 http://deb.debian.org/debian trixie-updates InRelease Reading package lists... Done Reading package lists... Done Building dependency tree... Done Reading state information... Done Calculating upgrade... Done The following packages will be upgraded: bind9-dnsutils bind9-host bind9-libs 3 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. Need to get 1,459 kB of archives. After this operation, 24.6 kB of additional disk space will be used. Get:1 http://security.debian.org trixie-security/main amd64 bind9-host amd64 1:9.20.15-1~deb13u1 [55.4 kB] Get:2 http://security.debian.org trixie-security/main amd64 bind9-dnsutils amd64 1:9.20.15-1~deb13u1 [165 kB] Get:3 http://security.debian.org trixie-security/main amd64 bind9-libs amd64 1:9.20.15-1~deb13u1 [1,238 kB] Fetched 1,459 kB in 1s (1,168 kB/s) apt-listchanges: Reading changelogs... (Reading database ... 20272 files and directories currently installed.) Preparing to unpack .../bind9-host_1%3a9.20.15-1~deb13u1_amd64.deb ... Unpacking bind9-host (1:9.20.15-1~deb13u1) over (1:9.20.11-4) ... Preparing to unpack .../bind9-dnsutils_1%3a9.20.15-1~deb13u1_amd64.deb ... Unpacking bind9-dnsutils (1:9.20.15-1~deb13u1) over (1:9.20.11-4) ... Preparing to unpack .../bind9-libs_1%3a9.20.15-1~deb13u1_amd64.deb ... Unpacking bind9-libs:amd64 (1:9.20.15-1~deb13u1) over (1:9.20.11-4) ... Setting up bind9-libs:amd64 (1:9.20.15-1~deb13u1) ... Setting up bind9-host (1:9.20.15-1~deb13u1) ... Setting up bind9-dnsutils (1:9.20.15-1~deb13u1) ... Processing triggers for man-db (2.13.1-1) ... Processing triggers for libc-bin (2.41-12) ... ✔️ Updated Container OS Reading package lists... Dones (Patience) Building dependency tree... Done Reading state information... Done Note, selecting 'tcl' instead of 'tclsh' Package libatlas-base-dev is not available, but is referred to by another package. This may mean that the package is missing, has been obsoleted, or is only available from another source Package python3-distutils is not available, but is referred to by another package. This may mean that the package is missing, has been obsoleted, or is only available from another source E: Package 'python3-distutils' has no installation candidate E: Unable to locate package python3.11 E: Couldn't find any package by glob 'python3.11' E: Couldn't find any package by regex 'python3.11' E: Unable to locate package python3.11-dev E: Couldn't find any package by glob 'python3.11-dev' E: Couldn't find any package by regex 'python3.11-dev' E: Package 'libatlas-base-dev' has no installation candidate [ERROR] in line 18: exit code 18 (Unknown error): while executing command apt-get install -y {jq,wget,xz-utils,python3,python3-dev,python3-distutils,gcc,pkg-config,libhdf5-dev,unzip,build-essential,automake,libtool,ccache,libusb-1.0-0-dev,apt-transport-https,python3.11,python3.11-dev,cmake,git,libgtk-3-dev,libavcodec-dev,libavformat-dev,libswscale-dev,libv4l-dev,libxvidcore-dev,libx264-dev,libjpeg-dev,libpng-dev,libtiff-dev,gfortran,openexr,libatlas-base-dev,libssl-dev,libtbbmalloc2,libtbb-dev,libdc1394-dev,libopenexr-dev,libgstreamer-plugins-base1.0-dev,libgstreamer1.0-dev,tclsh,libopenblas-dev,liblapack-dev,make,moreutils} ```

kerem commented

Author

Owner

@ShaMAD commented on GitHub (Nov 1, 2025):

Sory, I have an incorrect configuration in config.yml, the correct settings are:

detectors:
  coral1:
    type: edgetpu
    device: pci:0
  coral2:
    type: edgetpu
    device: pci:1
model:
  path: /edgetpu_model.tflite
  labelmap_path: /labelmap.txt

@ShaMAD commented on GitHub (Nov 1, 2025): Sory, I have an incorrect configuration in config.yml, the correct settings are: ``` detectors: coral1: type: edgetpu device: pci:0 coral2: type: edgetpu device: pci:1 model: path: /edgetpu_model.tflite labelmap_path: /labelmap.txt ```

kerem commented

Author

Owner

@jimisola commented on GitHub (Nov 1, 2025):

Just to confirm, Frigate LXC support was removed with https://github.com/community-scripts/ProxmoxVE/pull/7972 and this issue (https://github.com/community-scripts/ProxmoxVE/issues/6915) is for 0.16 to be supported again. But, it is not stable yet for release?

I'm going to use it with Ryzen 3 5300U and Coral TPU USB (awaiting shipment on the TPU). Is there anything I can do to help out testing?

The url to use it then this bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVED/main/ct/frigate.sh)" or is there a branch for it?

Is there a reason why the the LXC runs as privileged when it seems to work as unprivileged as well https://github.com/blakeblackshear/frigate/discussions/5773?

@jimisola commented on GitHub (Nov 1, 2025): Just to confirm, Frigate LXC support was removed with https://github.com/community-scripts/ProxmoxVE/pull/7972 and this issue (https://github.com/community-scripts/ProxmoxVE/issues/6915) is for 0.16 to be supported again. But, it is not stable yet for release? I'm going to use it with Ryzen 3 5300U and Coral TPU USB (awaiting shipment on the TPU). Is there anything I can do to help out testing? The url to use it then this `bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVED/main/ct/frigate.sh)"` or is there a branch for it? Is there a reason why the the LXC runs as privileged when it seems to work as unprivileged as well https://github.com/blakeblackshear/frigate/discussions/5773?

kerem commented

Author

Owner

@michelroegl-brunner commented on GitHub (Nov 2, 2025):

Correct, at this point there is no stable 0.16. install script. ProxmoxVED is the development repository, wich the scripts can change quite a bit without notice.

@michelroegl-brunner commented on GitHub (Nov 2, 2025): Correct, at this point there is no stable 0.16. install script. ProxmoxVED is the development repository, wich the scripts can change quite a bit without notice.

kerem commented

Author

Owner

@jimisola commented on GitHub (Nov 2, 2025):

Ok. Went with a manual unprivileged installation similar to this instruction. Works well with iGPU etc.

@jimisola commented on GitHub (Nov 2, 2025): Ok. Went with a manual **_unprivileged_** installation similar to this [instruction](https://www.mostlychris.com/installing-frigate-nvr-on-proxmox-in-an-lxc-container/). Works well with iGPU etc.

kerem commented

Author

Owner

@LeidenSpain commented on GitHub (Nov 2, 2025):

Hello,

I know this is a development version, so I'm sharing my error to provide more information.

I ran the script mentioned above, and it gave me this error without me doing anything.
I'm copying the entire installation code, as I'm not sure what information might be useful to you.

    ______     _             __     
   / ____/____(_)___ _____ _/ /____ 
  / /_  / ___/ / __ `/ __ `/ __/ _ \
 / __/ / /  / / /_/ / /_/ / /_/  __/
/_/   /_/  /_/\__, /\__,_/\__/\___/ 
             /____/                 
  🧩  Using Advanced Install on node PVE
  💡  PVE Version 9.0.11 (Kernel: 6.14.11-4-pve)
  💡  PVE Version 9.0.11 (Kernel: 6.14.11-4-pve)
  🖥️  Operating System: debian
  🌟  Version: 13
  📦  Container Type: Privileged
  🔐  Root Password: ********
  🆔  Container ID: 114
  🏠  Hostname: frigate
  💾  Disk Size: 20 GB
  🧠  CPU Cores: 4
  🛠️  RAM Size: 4096 MiB
  🌉  Bridge: vmbr0
  📡  IPv4 Address: 192.168.8.14/24
  🌐  Gateway IP Address: 192.168.8.1
  📡  IPv6: Disabled
  📡  APT-Cacher IP Address: Default
  ⚙️  Interface MTU Size: Default
  🔍  DNS Search Domain: Host
  📡  DNS Server IP Address: Host
  🏷️  Vlan: Default
  📡  Tags: nvr
  🔑  Root SSH Access: no
  🗂️  Enable FUSE Support: no
  🔍  Verbose Mode: yes
  🚀  Creating a Frigate LXC using the above advanced settings
  ✔️   Updated var_template_storage → local
  ✔️   Updated var_container_storage → local-lvm
  ✔️   Storage configuration saved to default.vars
  ✔️   Storage local (Free: 76.2MB  Used: 13.0MB) [Template]
  ✔️   Storage local-lvm (Free: 704.5MB  Used: 89.9MB) [Container]
  ⏳   Searching for template 'debian-13'  - debian-13-standard_13.1-2_amd64.tar.zst
  ✔️   Template debian-13-standard_13.1-2_amd64.tar.zst [online]
  ✔️   LXC Container 114 was successfully created.
  ✔️   USB passthrough configured
  ✔️   INTEL GPU passthrough configured (2 devices)
  ✔️   Started LXC Container
   💡   Network reachable but gateway check failed
  ✔️   Device GIDs updated successfully
  ✔️   Customized LXC Container
  ✔️   Set up Container OS
  ✔️   Network Connected: 192.168.8.14 
  ✔️   IPv4 Internet Connected
   ✖️   IPv6 Internet Not Connected
  ✔️   Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) git.community-scripts.org:(✔️ )
Hit:1 http://deb.debian.org/debian trixie InRelease
Hit:2 http://security.debian.org trixie-security InRelease
Hit:3 http://deb.debian.org/debian trixie-updates InRelease
Reading package lists... Done
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
Calculating upgrade... Done
The following packages will be upgraded:
  bind9-dnsutils bind9-host bind9-libs
3 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
Need to get 1,459 kB of archives.
After this operation, 24.6 kB of additional disk space will be used.
Get:1 http://security.debian.org trixie-security/main amd64 bind9-host amd64 1:9.20.15-1~deb13u1 [55.4 kB]
Get:2 http://security.debian.org trixie-security/main amd64 bind9-dnsutils amd64 1:9.20.15-1~deb13u1 [165 kB]
Get:3 http://security.debian.org trixie-security/main amd64 bind9-libs amd64 1:9.20.15-1~deb13u1 [1,238 kB]
Fetched 1,459 kB in 0s (5,887 kB/s)   
apt-listchanges: Reading changelogs...
(Reading database ... 20272 files and directories currently installed.)
Preparing to unpack .../bind9-host_1%3a9.20.15-1~deb13u1_amd64.deb ...
Unpacking bind9-host (1:9.20.15-1~deb13u1) over (1:9.20.11-4) ...
Preparing to unpack .../bind9-dnsutils_1%3a9.20.15-1~deb13u1_amd64.deb ...
Unpacking bind9-dnsutils (1:9.20.15-1~deb13u1) over (1:9.20.11-4) ...
Preparing to unpack .../bind9-libs_1%3a9.20.15-1~deb13u1_amd64.deb ...
Unpacking bind9-libs:amd64 (1:9.20.15-1~deb13u1) over (1:9.20.11-4) ...
Setting up bind9-libs:amd64 (1:9.20.15-1~deb13u1) ...
Setting up bind9-host (1:9.20.15-1~deb13u1) ...
Setting up bind9-dnsutils (1:9.20.15-1~deb13u1) ...
Processing triggers for man-db (2.13.1-1) ...
Processing triggers for libc-bin (2.41-12) ...
  ✔️   Updated Container OS
Reading package lists... Done (Patience)
Building dependency tree... Done
Reading state information... Done
Note, selecting 'tcl' instead of 'tclsh'
Package libatlas-base-dev is not available, but is referred to by another package.
This may mean that the package is missing, has been obsoleted, or
is only available from another source

Package python3-distutils is not available, but is referred to by another package.
This may mean that the package is missing, has been obsoleted, or
is only available from another source

E: Package 'python3-distutils' has no installation candidate
E: Unable to locate package python3.11
E: Couldn't find any package by glob 'python3.11'
E: Couldn't find any package by regex 'python3.11'
E: Unable to locate package python3.11-dev
E: Couldn't find any package by glob 'python3.11-dev'
E: Couldn't find any package by regex 'python3.11-dev'
E: Package 'libatlas-base-dev' has no installation candidate

[ERROR] in line 18: exit code 18 (Unknown error): while executing command  apt-get install -y {jq,wget,xz-utils,python3,python3-dev,python3-distutils,gcc,pkg-config,libhdf5-dev,unzip,build-essential,automake,libtool,ccache,libusb-1.0-0-dev,apt-transport-https,python3.11,python3.11-dev,cmake,git,libgtk-3-dev,libavcodec-dev,libavformat-dev,libswscale-dev,libv4l-dev,libxvidcore-dev,libx264-dev,libjpeg-dev,libpng-dev,libtiff-dev,gfortran,openexr,libatlas-base-dev,libssl-dev,libtbbmalloc2,libtbb-dev,libdc1394-dev,libopenexr-dev,libgstreamer-plugins-base1.0-dev,libgstreamer1.0-dev,tclsh,libopenblas-dev,liblapack-dev,make,moreutils}

I have a coral chip connected via USB.

@LeidenSpain commented on GitHub (Nov 2, 2025): Hello, I know this is a development version, so I'm sharing my error to provide more information. I ran the script mentioned above, and it gave me this error without me doing anything. I'm copying the entire installation code, as I'm not sure what information might be useful to you. ``` ______ _ __ / ____/____(_)___ _____ _/ /____ / /_ / ___/ / __ `/ __ `/ __/ _ \ / __/ / / / / /_/ / /_/ / /_/ __/ /_/ /_/ /_/\__, /\__,_/\__/\___/ /____/ 🧩 Using Advanced Install on node PVE 💡 PVE Version 9.0.11 (Kernel: 6.14.11-4-pve) 💡 PVE Version 9.0.11 (Kernel: 6.14.11-4-pve) 🖥️ Operating System: debian 🌟 Version: 13 📦 Container Type: Privileged 🔐 Root Password: ******** 🆔 Container ID: 114 🏠 Hostname: frigate 💾 Disk Size: 20 GB 🧠 CPU Cores: 4 🛠️ RAM Size: 4096 MiB 🌉 Bridge: vmbr0 📡 IPv4 Address: 192.168.8.14/24 🌐 Gateway IP Address: 192.168.8.1 📡 IPv6: Disabled 📡 APT-Cacher IP Address: Default ⚙️ Interface MTU Size: Default 🔍 DNS Search Domain: Host 📡 DNS Server IP Address: Host 🏷️ Vlan: Default 📡 Tags: nvr 🔑 Root SSH Access: no 🗂️ Enable FUSE Support: no 🔍 Verbose Mode: yes 🚀 Creating a Frigate LXC using the above advanced settings ✔️ Updated var_template_storage → local ✔️ Updated var_container_storage → local-lvm ✔️ Storage configuration saved to default.vars ✔️ Storage local (Free: 76.2MB Used: 13.0MB) [Template] ✔️ Storage local-lvm (Free: 704.5MB Used: 89.9MB) [Container] ⏳ Searching for template 'debian-13' - debian-13-standard_13.1-2_amd64.tar.zst ✔️ Template debian-13-standard_13.1-2_amd64.tar.zst [online] ✔️ LXC Container 114 was successfully created. ✔️ USB passthrough configured ✔️ INTEL GPU passthrough configured (2 devices) ✔️ Started LXC Container 💡 Network reachable but gateway check failed ✔️ Device GIDs updated successfully ✔️ Customized LXC Container ✔️ Set up Container OS ✔️ Network Connected: 192.168.8.14 ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) git.community-scripts.org:(✔️ ) Hit:1 http://deb.debian.org/debian trixie InRelease Hit:2 http://security.debian.org trixie-security InRelease Hit:3 http://deb.debian.org/debian trixie-updates InRelease Reading package lists... Done Reading package lists... Done Building dependency tree... Done Reading state information... Done Calculating upgrade... Done The following packages will be upgraded: bind9-dnsutils bind9-host bind9-libs 3 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. Need to get 1,459 kB of archives. After this operation, 24.6 kB of additional disk space will be used. Get:1 http://security.debian.org trixie-security/main amd64 bind9-host amd64 1:9.20.15-1~deb13u1 [55.4 kB] Get:2 http://security.debian.org trixie-security/main amd64 bind9-dnsutils amd64 1:9.20.15-1~deb13u1 [165 kB] Get:3 http://security.debian.org trixie-security/main amd64 bind9-libs amd64 1:9.20.15-1~deb13u1 [1,238 kB] Fetched 1,459 kB in 0s (5,887 kB/s) apt-listchanges: Reading changelogs... (Reading database ... 20272 files and directories currently installed.) Preparing to unpack .../bind9-host_1%3a9.20.15-1~deb13u1_amd64.deb ... Unpacking bind9-host (1:9.20.15-1~deb13u1) over (1:9.20.11-4) ... Preparing to unpack .../bind9-dnsutils_1%3a9.20.15-1~deb13u1_amd64.deb ... Unpacking bind9-dnsutils (1:9.20.15-1~deb13u1) over (1:9.20.11-4) ... Preparing to unpack .../bind9-libs_1%3a9.20.15-1~deb13u1_amd64.deb ... Unpacking bind9-libs:amd64 (1:9.20.15-1~deb13u1) over (1:9.20.11-4) ... Setting up bind9-libs:amd64 (1:9.20.15-1~deb13u1) ... Setting up bind9-host (1:9.20.15-1~deb13u1) ... Setting up bind9-dnsutils (1:9.20.15-1~deb13u1) ... Processing triggers for man-db (2.13.1-1) ... Processing triggers for libc-bin (2.41-12) ... ✔️ Updated Container OS Reading package lists... Done (Patience) Building dependency tree... Done Reading state information... Done Note, selecting 'tcl' instead of 'tclsh' Package libatlas-base-dev is not available, but is referred to by another package. This may mean that the package is missing, has been obsoleted, or is only available from another source Package python3-distutils is not available, but is referred to by another package. This may mean that the package is missing, has been obsoleted, or is only available from another source E: Package 'python3-distutils' has no installation candidate E: Unable to locate package python3.11 E: Couldn't find any package by glob 'python3.11' E: Couldn't find any package by regex 'python3.11' E: Unable to locate package python3.11-dev E: Couldn't find any package by glob 'python3.11-dev' E: Couldn't find any package by regex 'python3.11-dev' E: Package 'libatlas-base-dev' has no installation candidate [ERROR] in line 18: exit code 18 (Unknown error): while executing command apt-get install -y {jq,wget,xz-utils,python3,python3-dev,python3-distutils,gcc,pkg-config,libhdf5-dev,unzip,build-essential,automake,libtool,ccache,libusb-1.0-0-dev,apt-transport-https,python3.11,python3.11-dev,cmake,git,libgtk-3-dev,libavcodec-dev,libavformat-dev,libswscale-dev,libv4l-dev,libxvidcore-dev,libx264-dev,libjpeg-dev,libpng-dev,libtiff-dev,gfortran,openexr,libatlas-base-dev,libssl-dev,libtbbmalloc2,libtbb-dev,libdc1394-dev,libopenexr-dev,libgstreamer-plugins-base1.0-dev,libgstreamer1.0-dev,tclsh,libopenblas-dev,liblapack-dev,make,moreutils} ``` I have a coral chip connected via USB.

kerem commented

Author

Owner

@jairignaciogg commented on GitHub (Nov 7, 2025):

@LeidenSpain is the debian version should be 12

@jairignaciogg commented on GitHub (Nov 7, 2025): @LeidenSpain is the debian version should be 12

kerem commented

Author

Owner

@LeidenSpain commented on GitHub (Nov 7, 2025):

@LeidenSpain is the debian version should be 12

OK, I try again

@LeidenSpain commented on GitHub (Nov 7, 2025): > [@LeidenSpain](https://github.com/LeidenSpain) is the debian version should be 12 OK, I try again

kerem commented

Author

Owner

@LeidenSpain commented on GitHub (Nov 7, 2025):

@LeidenSpain is the debian version should be 12

I ran the script again and it didn't ask for the Debian version, so the script itself decides it should be 13.

@LeidenSpain commented on GitHub (Nov 7, 2025): > [@LeidenSpain](https://github.com/LeidenSpain) is the debian version should be 12 I ran the script again and it didn't ask for the Debian version, so the script itself decides it should be 13.

kerem commented

Author

Owner

@mapsindustrial commented on GitHub (Nov 7, 2025):

@LeidenSpain is the debian version should be 12

I ran the script again and it didn't ask for the Debian version, so the script itself decides it should be 13.

the same fails for me

@mapsindustrial commented on GitHub (Nov 7, 2025): > > [@LeidenSpain](https://github.com/LeidenSpain) is the debian version should be 12 > > I ran the script again and it didn't ask for the Debian version, so the script itself decides it should be 13. the same fails for me

kerem commented

Author

Owner

@jairignaciogg commented on GitHub (Nov 15, 2025):

It's because of the version of the script you're using. I'm guessing it's the one from community-scripts/ProxmoxVED, which has Debian 13 in ct/frigate.sh and build.func is different from the one in ProxmoxVE (all mentioned above).

I recommend using the script by @remz1337 :
bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"
Although it currently returns an error in "Installing Frigate Dependencies," it is an error in Frigate, and I believe it will be resolved in version 0.17, as mentioned above.

If you really want to try the script, you can try the fork I made to remz1337's version. I just removed nvidia-pyindex from requirements.txt.
bash -c "$(curl -fsSL https://raw.githubusercontent.com/jairignaciogg/ProxmoxVE/nvidiapyindexerror/ct/frigate.sh)"
(although I may remove it once v0.17 is released)

@jairignaciogg commented on GitHub (Nov 15, 2025): It's because of the version of the script you're using. I'm guessing it's the one from community-scripts/ProxmoxVED, which has Debian 13 in ct/frigate.sh and build.func is different from the one in ProxmoxVE (all mentioned above). I recommend using the script by @remz1337 : `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"` Although it currently returns an error in "Installing Frigate Dependencies," it is an error in Frigate, and I believe it will be resolved in version 0.17, as mentioned [above](https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3469443000). If you really want to try the script, you can try the fork I made to remz1337's version. I just removed nvidia-pyindex from requirements.txt. `bash -c "$(curl -fsSL https://raw.githubusercontent.com/jairignaciogg/ProxmoxVE/nvidiapyindexerror/ct/frigate.sh)"` (although I may remove it once v0.17 is released)

kerem commented

Author

Owner

@mapsindustrial commented on GitHub (Nov 15, 2025):

bash -c "$(curl -fsSL https://raw.githubusercontent.com/jairignaciogg/ProxmoxVE/nvidiapyindexerror/ct/frigate.sh)"

Does this LXC have Frigate installed? What is the port? NOT START FOR ME IN THE PORT 5000

What are the username and password for the console?

Thank you very much.

@mapsindustrial commented on GitHub (Nov 15, 2025): > bash -c "$(curl -fsSL https://raw.githubusercontent.com/jairignaciogg/ProxmoxVE/nvidiapyindexerror/ct/frigate.sh)" Does this LXC have Frigate installed? What is the port? NOT START FOR ME IN THE PORT 5000 What are the username and password for the console? Thank you very much.

kerem commented

Author

Owner

@stefan131 commented on GitHub (Nov 24, 2025):

I ran into the same issues as @mapsindustrial , the install fails which leaves the container without frigate installed. It also doesn't auto-login so you have to use pct enter {container_id}.

I did some debugging had some help from AI which leaded me to this version: https://github.com/stefan131/ProxmoxVE/raw/frigate-16-fix/ct/frigate.sh

I've tested this script on my HP EliteDesk with Intel i5-7500, no TPU, no NVIDIA.

Issues found and fixed

NVIDIA Passthrough Default: Changed default from "yes" to "no" since most users don't have NVIDIA GPUs. This prevented the installer from exiting when NVIDIA drivers weren't found on the host. It can be changed through advanced installer options.
nvidia-pyindex Installation Failure: The package fails to build on non-NVIDIA systems. Now conditionally removed based on GPU detection (credit to @jairignaciogg for
the original fix: github.com/jairignaciogg/ProxmoxVE@db7f04029d).
FFmpeg Missing: Added FFmpeg to dependencies and created required symlinks at /usr/lib/ffmpeg/7.0/bin/.
s6-overlay Shebangs: Replaced #!/command/with-contenv bash with #!/bin/bash for LXC compatibility.
TypeScript Build Errors: Skipped tsc type checking in the web build to avoid compilation errors.
Python Package Structure: Created missing __init__.py in frigate/object_detection/ package.
Hardware Acceleration: Disabled hwaccel_args for non-NVIDIA systems to prevent VAAPI errors on Intel iGPU in LXC.
Python Module Resolution: Added PYTHONPATH=/opt/frigate to Frigate systemd service.

Working Fork

Fork available at: https://github.com/stefan131/ProxmoxVE/tree/frigate-16-fix

Installation command:

bash -c "$(wget -qLO - https://github.com/stefan131/ProxmoxVE/raw/frigate-16-fix/ct/frigate.sh)"

Test Results

Output of installer

    ______     _             __     
   / ____/____(_)___ _____ _/ /____ 
  / /_  / ___/ / __ `/ __ `/ __/ _ \
 / __/ / /  / / /_/ / /_/ / /_/  __/
/_/   /_/  /_/\__, /\__,_/\__/\___/ 
             /____/                 
  ⚙️  Using Default Settings on node pvegitrunner1
  🆔  Container ID: {container_id}
  🖥️  Operating System: debian (12)
  📦  Container Type: Unprivileged
  💾  Disk Size: 50 GB
  🧠  CPU Cores: 4
  🛠️  RAM Size: 4096 MiB
  🧩  Add common sudo user with SSH access: no
  🧩  Enable Shared Mount: no
  🧩  Configure Postfix Satellite: no
  🧩  Passthrough Nvidia GPU: no
  🚀  Creating a Frigate LXC using the above default settings
  
  ✔️   Storage local (Free: xxGB  Used: xxGB) [Template]
  ✔️   Storage local-lvm (Free: xxGB  Used: xxGB) [Container]
  ✔️   Template debian-12-standard_12.12-1_amd64.tar.zst [local]
  ✔️   LXC Container 102 was successfully created.

 ⚙️  Configuring VAAPI passthrough for LXC container
 ⚠️  Container is unprivileged – VAAPI passthrough may not work without additional host configuration (e.g., idmap).
 ℹ️  VAAPI enables GPU hardware acceleration (e.g., for video transcoding in Jellyfin or Plex).

➤ Automatically mount all available VAAPI devices? [Y/n]: n
  ✔️   Started LXC Container
  ✔️   Network in LXC is reachable (ping)
  ✔️   Customized LXC Container
Customizing LXC creation
  ✔️   Rebooted LXC
  ✔️   Set up Container OS
  ✔️   Network Connected: {my_internal_ip}
  ✔️   IPv4 Internet Connected
   ✖️   IPv6 Internet Not Connected
  ✔️   Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
  ✔️   Updated Container OS
  ✔️   Set up Container
  ✔️   Installed APT proxy client
  ✔️   Installed sudo
  ✔️   Post install script completed.
  ✔️   Set up Container OS
  ✔️   Network Connected: {my_internal_ip}
  ✔️   IPv4 Internet Connected
   ✖️   IPv6 Internet Not Connected
  ✔️   Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
  ✔️   Updated Container OS
  ✔️   Installed Dependencies
  ✔️   Created FFmpeg symlinks
  ✔️   Set Up Hardware Acceleration
  ✔️   Setup environment
  ✔️   Deployed: frigate (0.16.2)
  ✔️   Downloaded Frigate source
  ✔️   Built Nginx
  ✔️   Built SQLite
  ✔️   Deployed: go2rtc (1.9.12)
  ✔️   Installed go2rtc
  ✔️   Installed Tempio
  ✔️   Built libUSB
  ✔️   Installed Pip
  ✔️   Installed Frigate Dependencies
  ✔️   Built pysqlite3
  ✔️   Setup Node.js 22
  ✔️   Installed Node.js modules: yarn
  ✔️   Installed NodeJS
  ✔️   Downloaded Coral TPU Model
  ✔️   Downloaded CPU Model
  ✔️   Built Audio Models
  ✔️   Built HailoRT
  ✔️   Installed OpenVino Runtime and Dev library
  ✔️   Downloaded OpenVino Model
  ✔️   Installed Frigate
  ✔️   AVX or SSE 4.2 Support Detected
  ✔️   Configured Openvino Object Detection Model
  ✔️   Created Services
  ✔️   Customized Container
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
  ✔️   Cleaned
Don't forget to edit the Frigate config file (/config/config.yml) and reboot. Example configuration at https://docs.frigate.video/configuration/
  ✔️   Set Container to Normal Resources
  ✔️   Completed Successfully!

  🚀  Frigate setup has been successfully initialized!
  💡   Access it using the following URL:
    🌐  http://{my_internal_id}:5000
root@elitedesk:~#

Tested on:

Hardware: HP EliteDesk, Intel i5-7500 (7th gen)
GPU: Intel HD Graphics 630 (integrated)
No TPU (software detection only)
Container: Proxmox LXC (Debian 12, unprivileged)

Credits

Original script: @remz1337
nvidia-pyindex fix: @jairignaciogg
Testing and additional fixes: @stefan131

This is just a quick fix, make sure to test thoroughly if planning to use in production environment

@stefan131 commented on GitHub (Nov 24, 2025): I ran into the same issues as @mapsindustrial , the install fails which leaves the container without frigate installed. It also doesn't auto-login so you have to use `pct enter {container_id}`. I did some debugging had some help from AI which leaded me to this version: `https://github.com/stefan131/ProxmoxVE/raw/frigate-16-fix/ct/frigate.sh` I've tested this script on my HP EliteDesk with Intel i5-7500, no TPU, no NVIDIA. ### Issues found and fixed 1. **NVIDIA Passthrough Default**: Changed default from "yes" to "no" since most users don't have NVIDIA GPUs. This prevented the installer from exiting when NVIDIA drivers weren't found on the host. It can be changed through advanced installer options. 2. **nvidia-pyindex Installation Failure**: The package fails to build on non-NVIDIA systems. Now conditionally removed based on GPU detection (credit to @jairignaciogg for the original fix: https://github.com/jairignaciogg/ProxmoxVE/commit/db7f04029d4fc3cbd96d8ad4e89d746544824200). 3. **FFmpeg Missing**: Added FFmpeg to dependencies and created required symlinks at `/usr/lib/ffmpeg/7.0/bin/`. 4. **s6-overlay Shebangs**: Replaced `#!/command/with-contenv bash` with `#!/bin/bash` for LXC compatibility. 5. **TypeScript Build Errors**: Skipped `tsc` type checking in the web build to avoid compilation errors. 6. **Python Package Structure**: Created missing `__init__.py` in `frigate/object_detection/` package. 7. **Hardware Acceleration**: Disabled `hwaccel_args` for non-NVIDIA systems to prevent VAAPI errors on Intel iGPU in LXC. 8. **Python Module Resolution**: Added `PYTHONPATH=/opt/frigate` to Frigate systemd service. ### Working Fork Fork available at: https://github.com/stefan131/ProxmoxVE/tree/frigate-16-fix **Installation command:** ```bash bash -c "$(wget -qLO - https://github.com/stefan131/ProxmoxVE/raw/frigate-16-fix/ct/frigate.sh)" ``` ### Test Results <details> <summary>Output of installer</summary> ``` ______ _ __ / ____/____(_)___ _____ _/ /____ / /_ / ___/ / __ `/ __ `/ __/ _ \ / __/ / / / / /_/ / /_/ / /_/ __/ /_/ /_/ /_/\__, /\__,_/\__/\___/ /____/ ⚙️ Using Default Settings on node pvegitrunner1 🆔 Container ID: {container_id} 🖥️ Operating System: debian (12) 📦 Container Type: Unprivileged 💾 Disk Size: 50 GB 🧠 CPU Cores: 4 🛠️ RAM Size: 4096 MiB 🧩 Add common sudo user with SSH access: no 🧩 Enable Shared Mount: no 🧩 Configure Postfix Satellite: no 🧩 Passthrough Nvidia GPU: no 🚀 Creating a Frigate LXC using the above default settings ✔️ Storage local (Free: xxGB Used: xxGB) [Template] ✔️ Storage local-lvm (Free: xxGB Used: xxGB) [Container] ✔️ Template debian-12-standard_12.12-1_amd64.tar.zst [local] ✔️ LXC Container 102 was successfully created. ⚙️ Configuring VAAPI passthrough for LXC container ⚠️ Container is unprivileged – VAAPI passthrough may not work without additional host configuration (e.g., idmap). ℹ️ VAAPI enables GPU hardware acceleration (e.g., for video transcoding in Jellyfin or Plex). ➤ Automatically mount all available VAAPI devices? [Y/n]: n ✔️ Started LXC Container ✔️ Network in LXC is reachable (ping) ✔️ Customized LXC Container Customizing LXC creation ✔️ Rebooted LXC ✔️ Set up Container OS ✔️ Network Connected: {my_internal_ip} ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) ✔️ Updated Container OS ✔️ Set up Container ✔️ Installed APT proxy client ✔️ Installed sudo ✔️ Post install script completed. ✔️ Set up Container OS ✔️ Network Connected: {my_internal_ip} ✔️ IPv4 Internet Connected ✖️ IPv6 Internet Not Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) ✔️ Updated Container OS ✔️ Installed Dependencies ✔️ Created FFmpeg symlinks ✔️ Set Up Hardware Acceleration ✔️ Setup environment ✔️ Deployed: frigate (0.16.2) ✔️ Downloaded Frigate source ✔️ Built Nginx ✔️ Built SQLite ✔️ Deployed: go2rtc (1.9.12) ✔️ Installed go2rtc ✔️ Installed Tempio ✔️ Built libUSB ✔️ Installed Pip ✔️ Installed Frigate Dependencies ✔️ Built pysqlite3 ✔️ Setup Node.js 22 ✔️ Installed Node.js modules: yarn ✔️ Installed NodeJS ✔️ Downloaded Coral TPU Model ✔️ Downloaded CPU Model ✔️ Built Audio Models ✔️ Built HailoRT ✔️ Installed OpenVino Runtime and Dev library ✔️ Downloaded OpenVino Model ✔️ Installed Frigate ✔️ AVX or SSE 4.2 Support Detected ✔️ Configured Openvino Object Detection Model ✔️ Created Services ✔️ Customized Container Reading package lists... Done Building dependency tree... Done Reading state information... Done 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. Reading package lists... Done Building dependency tree... Done Reading state information... Done ✔️ Cleaned Don't forget to edit the Frigate config file (/config/config.yml) and reboot. Example configuration at https://docs.frigate.video/configuration/ ✔️ Set Container to Normal Resources ✔️ Completed Successfully! 🚀 Frigate setup has been successfully initialized! 💡 Access it using the following URL: 🌐 http://{my_internal_id}:5000 root@elitedesk:~# ``` </details> Tested on: - Hardware: HP EliteDesk, Intel i5-7500 (7th gen) - GPU: Intel HD Graphics 630 (integrated) - No TPU (software detection only) - Container: Proxmox LXC (Debian 12, unprivileged) Credits - Original script: @remz1337 - nvidia-pyindex fix: @jairignaciogg - Testing and additional fixes: @stefan131 This is just a quick fix, make sure to test thoroughly if planning to use in production environment

kerem commented

Author

Owner

@jairignaciogg commented on GitHub (Nov 25, 2025):

@stefan131 It would be great if you could add this change to your fork. dbdfef5 to support proxmox 9.1

@jairignaciogg commented on GitHub (Nov 25, 2025): @stefan131 It would be great if you could add this change to your fork. [dbdfef5](https://github.com/community-scripts/ProxmoxVE/pull/9280/commits/dbdfef5c6ebf473ad895b6d458de9d57fce340e7) to support proxmox 9.1

kerem commented

Author

Owner

@mapsindustrial commented on GitHub (Nov 25, 2025):

I ran into the same issues as @mapsindustrial , the install fails which leaves the container without frigate installed. It also doesn't auto-login so you have to use pct enter {container_id}.

I did some debugging had some help from AI which leaded me to this version: https://github.com/stefan131/ProxmoxVE/raw/frigate-16-fix/ct/frigate.sh

I've tested this script on my HP EliteDesk with Intel i5-7500, no TPU, no NVIDIA.

Issues found and fixed

NVIDIA Passthrough Default: Changed default from "yes" to "no" since most users don't have NVIDIA GPUs. This prevented the installer from exiting when NVIDIA drivers weren't found on the host. It can be changed through advanced installer options.

nvidia-pyindex Installation Failure: The package fails to build on non-NVIDIA systems. Now conditionally removed based on GPU detection (credit to @jairignaciogg for
the original fix: jairignaciogg@db7f040).

FFmpeg Missing: Added FFmpeg to dependencies and created required symlinks at /usr/lib/ffmpeg/7.0/bin/.

s6-overlay Shebangs: Replaced #!/command/with-contenv bash with #!/bin/bash for LXC compatibility.

TypeScript Build Errors: Skipped tsc type checking in the web build to avoid compilation errors.

Python Package Structure: Created missing __init__.py in frigate/object_detection/ package.

Hardware Acceleration: Disabled hwaccel_args for non-NVIDIA systems to prevent VAAPI errors on Intel iGPU in LXC.

Python Module Resolution: Added PYTHONPATH=/opt/frigate to Frigate systemd service.

Working Fork

Fork available at: https://github.com/stefan131/ProxmoxVE/tree/frigate-16-fix

Installation command:

bash -c "$(wget -qLO - https://github.com/stefan131/ProxmoxVE/raw/frigate-16-fix/ct/frigate.sh)"

Test Results

Output of installer
    ______     _             __     
   / ____/____(_)___ _____ _/ /____ 
  / /_  / ___/ / __ `/ __ `/ __/ _ \
 / __/ / /  / / /_/ / /_/ / /_/  __/
/_/   /_/  /_/\__, /\__,_/\__/\___/ 
             /____/                 
  ⚙️  Using Default Settings on node pvegitrunner1
  🆔  Container ID: {container_id}
  🖥️  Operating System: debian (12)
  📦  Container Type: Unprivileged
  💾  Disk Size: 50 GB
  🧠  CPU Cores: 4
  🛠️  RAM Size: 4096 MiB
  🧩  Add common sudo user with SSH access: no
  🧩  Enable Shared Mount: no
  🧩  Configure Postfix Satellite: no
  🧩  Passthrough Nvidia GPU: no
  🚀  Creating a Frigate LXC using the above default settings
  
  ✔️   Storage local (Free: xxGB  Used: xxGB) [Template]
  ✔️   Storage local-lvm (Free: xxGB  Used: xxGB) [Container]
  ✔️   Template debian-12-standard_12.12-1_amd64.tar.zst [local]
  ✔️   LXC Container 102 was successfully created.

 ⚙️  Configuring VAAPI passthrough for LXC container
 ⚠️  Container is unprivileged – VAAPI passthrough may not work without additional host configuration (e.g., idmap).
 ℹ️  VAAPI enables GPU hardware acceleration (e.g., for video transcoding in Jellyfin or Plex).

➤ Automatically mount all available VAAPI devices? [Y/n]: n
  ✔️   Started LXC Container
  ✔️   Network in LXC is reachable (ping)
  ✔️   Customized LXC Container
Customizing LXC creation
  ✔️   Rebooted LXC
  ✔️   Set up Container OS
  ✔️   Network Connected: {my_internal_ip}
  ✔️   IPv4 Internet Connected
   ✖️   IPv6 Internet Not Connected
  ✔️   Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
  ✔️   Updated Container OS
  ✔️   Set up Container
  ✔️   Installed APT proxy client
  ✔️   Installed sudo
  ✔️   Post install script completed.
  ✔️   Set up Container OS
  ✔️   Network Connected: {my_internal_ip}
  ✔️   IPv4 Internet Connected
   ✖️   IPv6 Internet Not Connected
  ✔️   Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
  ✔️   Updated Container OS
  ✔️   Installed Dependencies
  ✔️   Created FFmpeg symlinks
  ✔️   Set Up Hardware Acceleration
  ✔️   Setup environment
  ✔️   Deployed: frigate (0.16.2)
  ✔️   Downloaded Frigate source
  ✔️   Built Nginx
  ✔️   Built SQLite
  ✔️   Deployed: go2rtc (1.9.12)
  ✔️   Installed go2rtc
  ✔️   Installed Tempio
  ✔️   Built libUSB
  ✔️   Installed Pip
  ✔️   Installed Frigate Dependencies
  ✔️   Built pysqlite3
  ✔️   Setup Node.js 22
  ✔️   Installed Node.js modules: yarn
  ✔️   Installed NodeJS
  ✔️   Downloaded Coral TPU Model
  ✔️   Downloaded CPU Model
  ✔️   Built Audio Models
  ✔️   Built HailoRT
  ✔️   Installed OpenVino Runtime and Dev library
  ✔️   Downloaded OpenVino Model
  ✔️   Installed Frigate
  ✔️   AVX or SSE 4.2 Support Detected
  ✔️   Configured Openvino Object Detection Model
  ✔️   Created Services
  ✔️   Customized Container
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
  ✔️   Cleaned
Don't forget to edit the Frigate config file (/config/config.yml) and reboot. Example configuration at https://docs.frigate.video/configuration/
  ✔️   Set Container to Normal Resources
  ✔️   Completed Successfully!

  🚀  Frigate setup has been successfully initialized!
  💡   Access it using the following URL:
    🌐  http://{my_internal_id}:5000
root@elitedesk:~# 
Tested on:

Hardware: HP EliteDesk, Intel i5-7500 (7th gen)

GPU: Intel HD Graphics 630 (integrated)

No TPU (software detection only)

Container: Proxmox LXC (Debian 12, unprivileged)

Credits

Original script: @remz1337

nvidia-pyindex fix: @jairignaciogg

Testing and additional fixes: @stefan131

This is just a quick fix, make sure to test thoroughly if planning to use in production environment

Great WORK !

But When I add the CORAL, it stops working for me :(

@mapsindustrial commented on GitHub (Nov 25, 2025): > I ran into the same issues as [@mapsindustrial](https://github.com/mapsindustrial) , the install fails which leaves the container without frigate installed. It also doesn't auto-login so you have to use `pct enter {container_id}`. > > I did some debugging had some help from AI which leaded me to this version: `https://github.com/stefan131/ProxmoxVE/raw/frigate-16-fix/ct/frigate.sh` > > I've tested this script on my HP EliteDesk with Intel i5-7500, no TPU, no NVIDIA. > > ### Issues found and fixed > 1. **NVIDIA Passthrough Default**: Changed default from "yes" to "no" since most users don't have NVIDIA GPUs. This prevented the installer from exiting when NVIDIA drivers weren't found on the host. It can be changed through advanced installer options. > 2. **nvidia-pyindex Installation Failure**: The package fails to build on non-NVIDIA systems. Now conditionally removed based on GPU detection (credit to [@jairignaciogg](https://github.com/jairignaciogg) for > the original fix: [jairignaciogg@db7f040](https://github.com/jairignaciogg/ProxmoxVE/commit/db7f04029d4fc3cbd96d8ad4e89d746544824200)). > 3. **FFmpeg Missing**: Added FFmpeg to dependencies and created required symlinks at `/usr/lib/ffmpeg/7.0/bin/`. > 4. **s6-overlay Shebangs**: Replaced `#!/command/with-contenv bash` with `#!/bin/bash` for LXC compatibility. > 5. **TypeScript Build Errors**: Skipped `tsc` type checking in the web build to avoid compilation errors. > 6. **Python Package Structure**: Created missing `__init__.py` in `frigate/object_detection/` package. > 7. **Hardware Acceleration**: Disabled `hwaccel_args` for non-NVIDIA systems to prevent VAAPI errors on Intel iGPU in LXC. > 8. **Python Module Resolution**: Added `PYTHONPATH=/opt/frigate` to Frigate systemd service. > > ### Working Fork > Fork available at: https://github.com/stefan131/ProxmoxVE/tree/frigate-16-fix > > **Installation command:** > > bash -c "$(wget -qLO - https://github.com/stefan131/ProxmoxVE/raw/frigate-16-fix/ct/frigate.sh)" > ### Test Results > Output of installer > ``` > ______ _ __ > / ____/____(_)___ _____ _/ /____ > / /_ / ___/ / __ `/ __ `/ __/ _ \ > / __/ / / / / /_/ / /_/ / /_/ __/ > /_/ /_/ /_/\__, /\__,_/\__/\___/ > /____/ > ⚙️ Using Default Settings on node pvegitrunner1 > 🆔 Container ID: {container_id} > 🖥️ Operating System: debian (12) > 📦 Container Type: Unprivileged > 💾 Disk Size: 50 GB > 🧠 CPU Cores: 4 > 🛠️ RAM Size: 4096 MiB > 🧩 Add common sudo user with SSH access: no > 🧩 Enable Shared Mount: no > 🧩 Configure Postfix Satellite: no > 🧩 Passthrough Nvidia GPU: no > 🚀 Creating a Frigate LXC using the above default settings > > ✔️ Storage local (Free: xxGB Used: xxGB) [Template] > ✔️ Storage local-lvm (Free: xxGB Used: xxGB) [Container] > ✔️ Template debian-12-standard_12.12-1_amd64.tar.zst [local] > ✔️ LXC Container 102 was successfully created. > > ⚙️ Configuring VAAPI passthrough for LXC container > ⚠️ Container is unprivileged – VAAPI passthrough may not work without additional host configuration (e.g., idmap). > ℹ️ VAAPI enables GPU hardware acceleration (e.g., for video transcoding in Jellyfin or Plex). > > ➤ Automatically mount all available VAAPI devices? [Y/n]: n > ✔️ Started LXC Container > ✔️ Network in LXC is reachable (ping) > ✔️ Customized LXC Container > Customizing LXC creation > ✔️ Rebooted LXC > ✔️ Set up Container OS > ✔️ Network Connected: {my_internal_ip} > ✔️ IPv4 Internet Connected > ✖️ IPv6 Internet Not Connected > ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) > ✔️ Updated Container OS > ✔️ Set up Container > ✔️ Installed APT proxy client > ✔️ Installed sudo > ✔️ Post install script completed. > ✔️ Set up Container OS > ✔️ Network Connected: {my_internal_ip} > ✔️ IPv4 Internet Connected > ✖️ IPv6 Internet Not Connected > ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) > ✔️ Updated Container OS > ✔️ Installed Dependencies > ✔️ Created FFmpeg symlinks > ✔️ Set Up Hardware Acceleration > ✔️ Setup environment > ✔️ Deployed: frigate (0.16.2) > ✔️ Downloaded Frigate source > ✔️ Built Nginx > ✔️ Built SQLite > ✔️ Deployed: go2rtc (1.9.12) > ✔️ Installed go2rtc > ✔️ Installed Tempio > ✔️ Built libUSB > ✔️ Installed Pip > ✔️ Installed Frigate Dependencies > ✔️ Built pysqlite3 > ✔️ Setup Node.js 22 > ✔️ Installed Node.js modules: yarn > ✔️ Installed NodeJS > ✔️ Downloaded Coral TPU Model > ✔️ Downloaded CPU Model > ✔️ Built Audio Models > ✔️ Built HailoRT > ✔️ Installed OpenVino Runtime and Dev library > ✔️ Downloaded OpenVino Model > ✔️ Installed Frigate > ✔️ AVX or SSE 4.2 Support Detected > ✔️ Configured Openvino Object Detection Model > ✔️ Created Services > ✔️ Customized Container > Reading package lists... Done > Building dependency tree... Done > Reading state information... Done > 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. > Reading package lists... Done > Building dependency tree... Done > Reading state information... Done > ✔️ Cleaned > Don't forget to edit the Frigate config file (/config/config.yml) and reboot. Example configuration at https://docs.frigate.video/configuration/ > ✔️ Set Container to Normal Resources > ✔️ Completed Successfully! > > 🚀 Frigate setup has been successfully initialized! > 💡 Access it using the following URL: > 🌐 http://{my_internal_id}:5000 > root@elitedesk:~# > ``` > > Tested on: > > * Hardware: HP EliteDesk, Intel i5-7500 (7th gen) > * GPU: Intel HD Graphics 630 (integrated) > * No TPU (software detection only) > * Container: Proxmox LXC (Debian 12, unprivileged) > > Credits > > * Original script: [@remz1337](https://github.com/remz1337) > * nvidia-pyindex fix: [@jairignaciogg](https://github.com/jairignaciogg) > * Testing and additional fixes: [@stefan131](https://github.com/stefan131) > > This is just a quick fix, make sure to test thoroughly if planning to use in production environment Great WORK ! But When I add the CORAL, it stops working for me :(

kerem commented

Author

Owner

@Pehun commented on GitHub (Nov 27, 2025):

launched on proxmox 8.2.4, intel CPU, no IGPU, Intel GPU, getting this error :
/dev/fd/63: line 287: wait_for: No record of process 22728
✔️ Setup environment
⠙ Downloading Frigate sourcecurl: (23) Failed writing body

@Pehun commented on GitHub (Nov 27, 2025): launched on proxmox 8.2.4, intel CPU, no IGPU, Intel GPU, getting this error : /dev/fd/63: line 287: wait_for: No record of process 22728 ✔️ Setup environment ⠙ Downloading Frigate sourcecurl: (23) Failed writing body

kerem commented

Author

Owner

@Pehun commented on GitHub (Nov 27, 2025):

once started, in logs i'm getting :
Unable to poll intel GPU stats: Failed to initialize PMU! (Permission denied)

@Pehun commented on GitHub (Nov 27, 2025): once started, in logs i'm getting : Unable to poll intel GPU stats: Failed to initialize PMU! (Permission denied)

kerem commented

Author

Owner

@stefan131 commented on GitHub (Nov 27, 2025):

@jairignaciogg thanks for the hint, I rebased on the original upstream so it should have all the latest changes now. I made a new branch on my fork frigate-fixes-clean. I'm planning to test if this version still works as expected this weekend, hopefully after fixing the issues below when I get more context

@mapsindustrial Good to hear it works! Can you share what happens when you add the coral? What type is it (USB/PCIe/M.2), and any error messages? I don't have one to test with unfortunately.

@Pehun hmm that's strange, can you share:

Your hardware setup (Intel Arc GPU or typo?)
Did the install finish or stop at that error?
Can you access the web ui of frigate?
Was the previous working fine without any errors?

@stefan131 commented on GitHub (Nov 27, 2025): @jairignaciogg thanks for the hint, I rebased on the original upstream so it should have all the latest changes now. I made a new branch on my fork `frigate-fixes-clean`. I'm planning to test if this version still works as expected this weekend, hopefully after fixing the issues below when I get more context @mapsindustrial Good to hear it works! Can you share what happens when you add the coral? What type is it (USB/PCIe/M.2), and any error messages? I don't have one to test with unfortunately. @Pehun hmm that's strange, can you share: - Your hardware setup (Intel Arc GPU or typo?) - Did the install finish or stop at that error? - Can you access the web ui of frigate? - Was the previous working fine without any errors?

kerem commented

Author

Owner

@Pehun commented on GitHub (Nov 28, 2025):

Hello,
I have an Intel ARC380 and an Intel Xeon E-2224 3.4Ghz
Gemini is telling me that my GPU is badly installed on proxmox.
Tinkering with him to fix it, I get the script working and making the LXC to work.
Despites hours spent, the only config to work is the following :
ffmpeg:
hwaccel_args:
- -hwaccel
- vaapi
# RETIRER : - -hwaccel_device /dev/dri/renderD128 (cause des conflits)

detectors:
openvino:
type: openvino
# REVENIR AU MODE AUTO
device: 'AUTO'

model:
width: 300
height: 300
input_tensor: nhwc
input_pixel_format: bgr
path: /openvino-model/ssdlite_mobilenet_v2.xml
labelmap_path: /openvino-model/coco_91cl_bkgr.txt

But the GPU is not used by frigate to detect... Gemini is telling me that frigate is not up to date enough to use the intel Arc...
I'm surprised because i'm using 16.2 image, on proxmox 8.2.4... if anybody had some magic advice to make the A380 to work...
I'm using it without problem on a plex lxc... video decoding seems to work, but not AI... on 3D render...
Here is the gpu top :
intel-gpu-top: Intel Dg2 (Gen12) @ /dev/dri/card0 - 697/1919 MHz; 47% RC6
483 irqs/s

  IMC reads:     4732 MiB/s
 IMC writes:     2024 MiB/s

     ENGINES     BUSY                                                 MI_SEMA MI_WAIT
   Render/3D    0.00% |                                             |      0%      0%
     Blitter    0.00% |                                             |      0%      0%
       Video    2.64% |█▎                                           |      0%      0%
VideoEnhance    2.41% |█▏                                           |      0%      0%
   [unknown]    0.00% |                                             |      0%      0%

PID NAME Render/3D Blitter Video VideoEnhance [unknown]
4389 ffmpeg | || ||▏ ||▏ || |
4383 ffmpeg | || ||▏ ||▏ || |
398570 ffmpeg | || || ||▏ || |
394872 ffmpeg | || ||▏ ||▏ || |
398561 ffmpeg | || ||▏ || || |
398594 ffmpeg | || ||▏ ||▏ || |
4395 ffmpeg | || ||▏ ||▏ || |
398606 ffmpeg | || ||▏ ||▏ || |
4406 ffmpeg | || ||▏ ||▏ || |
398613 ffmpeg | || ||▏ || || |
4409 ffmpeg | || ||▏ || || |
398383 frigate.detecto | || || || || |

@Pehun commented on GitHub (Nov 28, 2025): Hello, I have an Intel ARC380 and an Intel Xeon E-2224 3.4Ghz Gemini is telling me that my GPU is badly installed on proxmox. Tinkering with him to fix it, I get the script working and making the LXC to work. Despites hours spent, the only config to work is the following : ffmpeg: hwaccel_args: - -hwaccel - vaapi # RETIRER : - -hwaccel_device /dev/dri/renderD128 (cause des conflits) detectors: openvino: type: openvino # REVENIR AU MODE AUTO device: 'AUTO' model: width: 300 height: 300 input_tensor: nhwc input_pixel_format: bgr path: /openvino-model/ssdlite_mobilenet_v2.xml labelmap_path: /openvino-model/coco_91cl_bkgr.txt But the GPU is not used by frigate to detect... Gemini is telling me that frigate is not up to date enough to use the intel Arc... I'm surprised because i'm using 16.2 image, on proxmox 8.2.4... if anybody had some magic advice to make the A380 to work... I'm using it without problem on a plex lxc... video decoding seems to work, but not AI... on 3D render... Here is the gpu top : intel-gpu-top: Intel Dg2 (Gen12) @ /dev/dri/card0 - 697/1919 MHz; 47% RC6 483 irqs/s IMC reads: 4732 MiB/s IMC writes: 2024 MiB/s ENGINES BUSY MI_SEMA MI_WAIT Render/3D 0.00% | | 0% 0% Blitter 0.00% | | 0% 0% Video 2.64% |█▎ | 0% 0% VideoEnhance 2.41% |█▏ | 0% 0% [unknown] 0.00% | | 0% 0% PID NAME Render/3D Blitter Video VideoEnhance [unknown] 4389 ffmpeg | || ||▏ ||▏ || | 4383 ffmpeg | || ||▏ ||▏ || | 398570 ffmpeg | || || ||▏ || | 394872 ffmpeg | || ||▏ ||▏ || | 398561 ffmpeg | || ||▏ || || | 398594 ffmpeg | || ||▏ ||▏ || | 4395 ffmpeg | || ||▏ ||▏ || | 398606 ffmpeg | || ||▏ ||▏ || | 4406 ffmpeg | || ||▏ ||▏ || | 398613 ffmpeg | || ||▏ || || | 4409 ffmpeg | || ||▏ || || | 398383 frigate.detecto | || || || || |

kerem commented

Author

Owner

@Pehun commented on GitHub (Dec 2, 2025):

https://github.com/blakeblackshear/frigate/issues/12266
Maybe this is my solution?

@Pehun commented on GitHub (Dec 2, 2025): https://github.com/blakeblackshear/frigate/issues/12266 Maybe this is my solution?

kerem commented

Author

Owner

@fjleon1980 commented on GitHub (Dec 2, 2025):

i just got a N150 mini pc that came with proxmox preinstalled. i did all updates and now it's running 9.1.1. Then i tried installing with remz script

bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"

I chose priviledged yes and nvidia no, since i want to use igpu

I am getting erratic dns resolution and i don't understand why, since this doesn't happen on any of my other computers, using the same DNS.

⠋ Installing Frigate Dependencies
[ERROR] in line 159: exit code 0: while executing command "$@" > /dev/null 2>&1
/dev/fd/63: line 126: pop_var_context: head of shell_variables not a function context
/dev/fd/63: line 288: local: can only be used in a function
curl: (6) Could not resolve host: raw.githubusercontent.com
/dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context

Using Advanced Settings on node proxmox
🖥️ Operating System: debian
🌟 Version: 12
📦 Container Type: Privileged
🔐 Root Password: ********
🆔 Container ID: 100
🏠 Hostname: frigate
💾 Disk Size: 50 GB
🧠 CPU Cores: 2
🛠️ RAM Size: 4096 MiB
🌉 Bridge: vmbr0
📡 IPv4: DHCP
📡 IPv6: SLAAC/AUTO
📡 APT-Cacher IP Address: Default
⚙️ Interface MTU Size: Default
🔍 DNS Search Domain: Host
📡 DNS Server IP Address: Host
Enable Firewall: 0
🏷️ Vlan: Default
📡 Tags: nvr
🔑 Root SSH Access: yes
🗂️ Enable FUSE Support: no
Add common sudo user with SSH access: no
Enable Shared Mount: no
Configure Postfix Satellite: no
Nvidia GPU Passthrough: no
🔍 Verbose Mode: no
🚀 Creating a Frigate LXC using the above advanced settings
✔️ Storage local (Free: 5.1GB Used: 8.4GB) [Template]
✔️ Storage vm-lvmthin (Free: 436.7GB Used: 0.0B) [Container]
✔️ Template debian-12-standard_12.12-1_amd64.tar.zst [online]
💡 Template debian-12-standard_12.12-1_amd64.tar.zst is missing or corrupted. Re-downloading.
✔️ Template download successful.
✔️ LXC Container 100 was successfully created.

⚙️ Configuring VAAPI passthrough for LXC container
ℹ️ VAAPI enables GPU hardware acceleration (e.g., for video transcoding in Jellyfin or Plex).

➤ Automatically mount all available VAAPI devices? [Y/n]:
✔️ Started LXC Container
💡 No network in LXC yet (try 1/10) – waiting...
✔️ Network in LXC is reachable (ping)
✔️ Customized LXC Container
Customizing LXC creation
✔️ Rebooted LXC
✔️ Set up Container OS
✔️ Network Connected: <ip's>
✔️ IPv4 Internet Connected
✔️ IPv6 Internet Connected
✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
✔️ Updated Container OS
✔️ Set up Container
✔️ Installed APT proxy client
✔️ Installed sudo
✔️ Post install script completed.
✔️ Set up Container OS
✔️ Network Connected: <ip's>
✔️ IPv4 Internet Connected
✔️ IPv6 Internet Connected
✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ )
✔️ Updated Container OS
✔️ Installed Dependencies
✔️ Set Up Hardware Acceleration
✔️ Setup environment
✔️ Deployed: frigate (0.16.2)
✔️ Downloaded Frigate source
✔️ Built Nginx
✔️ Built SQLite
✔️ Deployed: go2rtc (1.9.12)
✔️ Installed go2rtc
✔️ Installed Tempio
✔️ Built libUSB
✔️ Installed Pip

I had a lot of issues trying to install container profiles due to dns resolution so i will keep trying

@fjleon1980 commented on GitHub (Dec 2, 2025): i just got a N150 mini pc that came with proxmox preinstalled. i did all updates and now it's running 9.1.1. Then i tried installing with remz script bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)" I chose priviledged yes and nvidia no, since i want to use igpu I am getting erratic dns resolution and i don't understand why, since this doesn't happen on any of my other computers, using the same DNS. ⠋ Installing Frigate Dependencies [ERROR] in line 159: exit code 0: while executing command "$@" > /dev/null 2>&1 /dev/fd/63: line 126: pop_var_context: head of shell_variables not a function context /dev/fd/63: line 288: local: can only be used in a function curl: (6) Could not resolve host: raw.githubusercontent.com /dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context Using Advanced Settings on node proxmox 🖥️ Operating System: debian 🌟 Version: 12 📦 Container Type: Privileged 🔐 Root Password: ******** 🆔 Container ID: 100 🏠 Hostname: frigate 💾 Disk Size: 50 GB 🧠 CPU Cores: 2 🛠️ RAM Size: 4096 MiB 🌉 Bridge: vmbr0 📡 IPv4: DHCP 📡 IPv6: SLAAC/AUTO 📡 APT-Cacher IP Address: Default ⚙️ Interface MTU Size: Default 🔍 DNS Search Domain: Host 📡 DNS Server IP Address: Host Enable Firewall: 0 🏷️ Vlan: Default 📡 Tags: nvr 🔑 Root SSH Access: yes 🗂️ Enable FUSE Support: no Add common sudo user with SSH access: no Enable Shared Mount: no Configure Postfix Satellite: no Nvidia GPU Passthrough: no 🔍 Verbose Mode: no 🚀 Creating a Frigate LXC using the above advanced settings ✔️ Storage local (Free: 5.1GB Used: 8.4GB) [Template] ✔️ Storage vm-lvmthin (Free: 436.7GB Used: 0.0B) [Container] ✔️ Template debian-12-standard_12.12-1_amd64.tar.zst [online] 💡 Template debian-12-standard_12.12-1_amd64.tar.zst is missing or corrupted. Re-downloading. ✔️ Template download successful. ✔️ LXC Container 100 was successfully created. ⚙️ Configuring VAAPI passthrough for LXC container ℹ️ VAAPI enables GPU hardware acceleration (e.g., for video transcoding in Jellyfin or Plex). ➤ Automatically mount all available VAAPI devices? [Y/n]: ✔️ Started LXC Container 💡 No network in LXC yet (try 1/10) – waiting... ✔️ Network in LXC is reachable (ping) ✔️ Customized LXC Container Customizing LXC creation ✔️ Rebooted LXC ✔️ Set up Container OS ✔️ Network Connected: <ip's> ✔️ IPv4 Internet Connected ✔️ IPv6 Internet Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) ✔️ Updated Container OS ✔️ Set up Container ✔️ Installed APT proxy client ✔️ Installed sudo ✔️ Post install script completed. ✔️ Set up Container OS ✔️ Network Connected: <ip's> ✔️ IPv4 Internet Connected ✔️ IPv6 Internet Connected ✔️ Git DNS: github.com:(✔️ ) raw.githubusercontent.com:(✔️ ) api.github.com:(✔️ ) ✔️ Updated Container OS ✔️ Installed Dependencies ✔️ Set Up Hardware Acceleration ✔️ Setup environment ✔️ Deployed: frigate (0.16.2) ✔️ Downloaded Frigate source ✔️ Built Nginx ✔️ Built SQLite ✔️ Deployed: go2rtc (1.9.12) ✔️ Installed go2rtc ✔️ Installed Tempio ✔️ Built libUSB ✔️ Installed Pip I had a lot of issues trying to install container profiles due to dns resolution so i will keep trying

kerem commented

Author

Owner

@fjleon1980 commented on GitHub (Dec 2, 2025):

ok so i tried again with verbose and it failed at the same issue posted here https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3469443000 so i guess i can't do anything but wait

@fjleon1980 commented on GitHub (Dec 2, 2025): ok so i tried again with verbose and it failed at the same issue posted here https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3469443000 so i guess i can't do anything but wait

kerem commented

Author

Owner

@fjleon1980 commented on GitHub (Dec 2, 2025):

@stefan131 any ETA of when your work will add proxmox 9.1 support? i can't use remz version due to the nvidia pyindex, which yours has, but remz supports 9.1

i downloaded the script to see if it's something that i could quickly change, but the script sources a file from the cloud, which i also downloaded but it doesn't reference nvidia-pyindex anywhere, which according to the fix it's something set in requirements.txt which i can't find, so don't know how to fix this myself

@fjleon1980 commented on GitHub (Dec 2, 2025): @stefan131 any ETA of when your work will add proxmox 9.1 support? i can't use remz version due to the nvidia pyindex, which yours has, but remz supports 9.1 i downloaded the script to see if it's something that i could quickly change, but the script sources a file from the cloud, which i also downloaded but it doesn't reference nvidia-pyindex anywhere, which according to the fix it's something set in requirements.txt which i can't find, so don't know how to fix this myself

kerem commented

Author

Owner

@hmronline commented on GitHub (Dec 2, 2025):

Tested on:
Hardware: Custom PC, Intel i7-10700F, 64 GB RAM
GPU: NVIDIA GeForce GTX 1650
TPU: None
Host: Proxmox 9.1.1
Container: Proxmox LXC (Debian 12, unprivileged)

Script:
bash -c "$(wget -qLO - https://github.com/stefan131/ProxmoxVE/raw/frigate-fixes-clean/ct/frigate.sh)"

Config:

root@pve:~# cat /opt/community-scripts/frigate.conf 
# frigate Configuration File
# Generated on Tue Dec  2 06:26:03 PM -03 2025

CT_TYPE="1"
DISK_SIZE="20"
CORE_COUNT="4"
RAM_SIZE="4096"
VERBOSE="yes"
PW="secret"
#CT_ID=100
HN="frigate"
BRG="vmbr0"
NET="dhcp"
IPV6_METHOD="none"
# Set this only if using "IPV6_METHOD=static"
#IPV6STATIC="fd00::1234/64"

GATE="none"
APT_CACHER_IP="none"
MTU="1500"
SD="none"
NS="none"
MAC="none"
VLAN="10"
SSH="yes"
SSH_AUTHORIZED_KEY=""
TAGS="community-script;nvr"
ENABLE_FUSE="no"
ENABLE_TUN="no"

Error (just the error part):

  ✔️   Updated Container OS
Reading package lists... Done (Patience)
Building dependency tree... Done
Reading state information... Done
Package libtbb2 is not available, but is referred to by another package.
This may mean that the package is missing, has been obsoleted, or
is only available from another source
However the following packages replace it:
  libtbbmalloc2

E: Package 'libtbb2' has no installation candidate
E: Unable to locate package libdc1394-22-dev

[ERROR] in line 18: exit code 0: while executing command $STD apt-get install -y {git,ca-certificates,automake,build-essential,xz-utils,libtool,ccache,pkg-config,libgtk-3-dev,libavcodec-dev,libavformat-dev,libswscale-dev,libv4l-dev,libxvidcore-dev,libx264-dev,libjpeg-dev,libpng-dev,libtiff-dev,gfortran,openexr,libatlas-base-dev,libssl-dev,libtbb2,libtbb-dev,libdc1394-22-dev,libopenexr-dev,libgstreamer-plugins-base1.0-dev,libgstreamer1.0-dev,gcc,gfortran,libopenblas-dev,liblapack-dev,libusb-1.0-0-dev,jq,moreutils}

[ERROR] in line 1361: exit code 0: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/install/${var_install}.sh)"

@hmronline commented on GitHub (Dec 2, 2025): **Tested on**: Hardware: Custom PC, Intel i7-10700F, 64 GB RAM GPU: NVIDIA GeForce GTX 1650 TPU: None Host: Proxmox 9.1.1 Container: Proxmox LXC (Debian 12, unprivileged) **Script**: `bash -c "$(wget -qLO - https://github.com/stefan131/ProxmoxVE/raw/frigate-fixes-clean/ct/frigate.sh)"` **Config**: ``` root@pve:~# cat /opt/community-scripts/frigate.conf # frigate Configuration File # Generated on Tue Dec 2 06:26:03 PM -03 2025 CT_TYPE="1" DISK_SIZE="20" CORE_COUNT="4" RAM_SIZE="4096" VERBOSE="yes" PW="secret" #CT_ID=100 HN="frigate" BRG="vmbr0" NET="dhcp" IPV6_METHOD="none" # Set this only if using "IPV6_METHOD=static" #IPV6STATIC="fd00::1234/64" GATE="none" APT_CACHER_IP="none" MTU="1500" SD="none" NS="none" MAC="none" VLAN="10" SSH="yes" SSH_AUTHORIZED_KEY="" TAGS="community-script;nvr" ENABLE_FUSE="no" ENABLE_TUN="no" ``` **Error** (just the error part): ``` ✔️ Updated Container OS Reading package lists... Done (Patience) Building dependency tree... Done Reading state information... Done Package libtbb2 is not available, but is referred to by another package. This may mean that the package is missing, has been obsoleted, or is only available from another source However the following packages replace it: libtbbmalloc2 E: Package 'libtbb2' has no installation candidate E: Unable to locate package libdc1394-22-dev [ERROR] in line 18: exit code 0: while executing command $STD apt-get install -y {git,ca-certificates,automake,build-essential,xz-utils,libtool,ccache,pkg-config,libgtk-3-dev,libavcodec-dev,libavformat-dev,libswscale-dev,libv4l-dev,libxvidcore-dev,libx264-dev,libjpeg-dev,libpng-dev,libtiff-dev,gfortran,openexr,libatlas-base-dev,libssl-dev,libtbb2,libtbb-dev,libdc1394-22-dev,libopenexr-dev,libgstreamer-plugins-base1.0-dev,libgstreamer1.0-dev,gcc,gfortran,libopenblas-dev,liblapack-dev,libusb-1.0-0-dev,jq,moreutils} [ERROR] in line 1361: exit code 0: while executing command lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/install/${var_install}.sh)" ```

kerem commented

Author

Owner

@hmronline commented on GitHub (Dec 2, 2025):

Tested on:
Hardware: Custom PC, Intel i7-10700F, 64 GB RAM
GPU: NVIDIA GeForce GTX 1650
TPU: None
Host: Proxmox 9.1.1
Container: Proxmox LXC (Debian 12, unprivileged)

Script:
bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"

Config:

root@pve:~# cat /opt/community-scripts/frigate.conf 
# frigate Configuration File
# Generated on Tue Dec  2 06:42:12 PM -03 2025

CT_TYPE="1"
DISK_SIZE="50"
CORE_COUNT="4"
RAM_SIZE="4096"
VERBOSE="yes"
PW="secret"
#CT_ID=100
HN="frigate"
BRG="vmbr0"
NET="dhcp"
IPV6_METHOD="none"

# Set this only if using "IPV6_METHOD=static"
#IPV6STATIC="fd00::1234/64"

GATE="none"
APT_CACHER_IP="none"
MTU="1500"
SD="none"
NS="none"
MAC="none"
VLAN="10"
SSH="no"
SSH_AUTHORIZED_KEY=""
TAGS="remz1337;nvr"
ENABLE_FUSE="no"
ENABLE_TUN="no"

Error (just the error part):

Building wheels for collected packages: nvidia-pyindex
  Building wheel for nvidia-pyindex (pyproject.toml) ... error
  error: subprocess-exited-with-error
  
  × Building wheel for nvidia-pyindex (pyproject.toml) did not run successfully.
  │ exit code: 1
  ╰─> [99 lines of output]
      /tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/dist.py:759: SetuptoolsDeprecationWarning: License classifiers are deprecated.
      !!
      
              ********************************************************************************
              Please consider removing the following classifiers in favor of a SPDX license expression:
      
              License :: OSI Approved :: Apache Software License
      
              See https://packaging.python.org/en/latest/guides/writing-pyproject-toml/#license for details.
              ********************************************************************************
      
      !!
        self._finalize_license_expression()
      running bdist_wheel
      running build
      running build_py
      creating build/lib/nvidia_pyindex
      copying nvidia_pyindex/cmdline.py -> build/lib/nvidia_pyindex
      copying nvidia_pyindex/__init__.py -> build/lib/nvidia_pyindex
      copying nvidia_pyindex/utils.py -> build/lib/nvidia_pyindex
      running egg_info
      writing nvidia_pyindex.egg-info/PKG-INFO
      writing dependency_links to nvidia_pyindex.egg-info/dependency_links.txt
      writing entry points to nvidia_pyindex.egg-info/entry_points.txt
      writing top-level names to nvidia_pyindex.egg-info/top_level.txt
      reading manifest file 'nvidia_pyindex.egg-info/SOURCES.txt'
      reading manifest template 'MANIFEST.in'
      adding license file 'LICENSE.md'
      writing manifest file 'nvidia_pyindex.egg-info/SOURCES.txt'
      installing to build/bdist.linux-x86_64/wheel
      running install
      %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
      COMMAND: InstallCommand
      %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
      Traceback (most recent call last):
        File "/tmp/pip-install-984lbo3f/nvidia-pyindex_3d7ae6811f234c6fbdeb80137ea7ec7c/nvidia_pyindex/utils.py", line 20, in get_configuration_files
          output = subprocess.check_output(command, stderr=subprocess.STDOUT)
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        File "/usr/lib/python3.11/subprocess.py", line 466, in check_output
          return run(*popenargs, stdout=PIPE, timeout=timeout, check=True,
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        File "/usr/lib/python3.11/subprocess.py", line 571, in run
          raise CalledProcessError(retcode, process.args,
      subprocess.CalledProcessError: Command '['/usr/bin/python3', '-c', 'from pip._internal.configuration import get_configuration_files; print(get_configuration_files())']' returned non-zero exit status 1.
      
      The above exception was the direct cause of the following exception:
      
      Traceback (most recent call last):
        File "/usr/local/lib/python3.11/dist-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 389, in <module>
          main()
        File "/usr/local/lib/python3.11/dist-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 373, in main
          json_out["return_val"] = hook(**hook_input["kwargs"])
                                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        File "/usr/local/lib/python3.11/dist-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 280, in build_wheel
          return _build_backend().build_wheel(
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/build_meta.py", line 435, in build_wheel
          return _build(['bdist_wheel', '--dist-info-dir', str(metadata_directory)])
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/build_meta.py", line 423, in _build
          return self._build_with_temp_dir(
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^
        File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/build_meta.py", line 404, in _build_with_temp_dir
          self.run_setup()
        File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/build_meta.py", line 512, in run_setup
          super().run_setup(setup_script=setup_script)
        File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/build_meta.py", line 317, in run_setup
          exec(code, locals())
        File "<string>", line 238, in <module>
        File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/__init__.py", line 115, in setup
          return distutils.core.setup(**attrs)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/_distutils/core.py", line 186, in setup
          return run_commands(dist)
                 ^^^^^^^^^^^^^^^^^^
        File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/_distutils/core.py", line 202, in run_commands
          dist.run_commands()
        File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/_distutils/dist.py", line 1002, in run_commands
          self.run_command(cmd)
        File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/dist.py", line 1102, in run_command
          super().run_command(command)
        File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/_distutils/dist.py", line 1021, in run_command
          cmd_obj.run()
        File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/command/bdist_wheel.py", line 405, in run
          self.run_command("install")
        File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/_distutils/cmd.py", line 357, in run_command
          self.distribution.run_command(command)
        File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/dist.py", line 1102, in run_command
          super().run_command(command)
        File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/_distutils/dist.py", line 1021, in run_command
          cmd_obj.run()
        File "<string>", line 172, in run
        File "<string>", line 71, in _install_nvidia_pypi_index
        File "/tmp/pip-install-984lbo3f/nvidia-pyindex_3d7ae6811f234c6fbdeb80137ea7ec7c/nvidia_pyindex/utils.py", line 24, in get_configuration_files
          raise RuntimeError(output) from e
      RuntimeError: Traceback (most recent call last):
        File "<string>", line 1, in <module>
      ModuleNotFoundError: No module named 'pip'
      
      [end of output]
  
  note: This error originates from a subprocess, and is likely not a problem with pip.
  ERROR: Failed building wheel for nvidia-pyindex
Failed to build nvidia-pyindex
error: failed-wheel-build-for-install

× Failed to build installable wheels for some pyproject.toml based projects
╰─> nvidia-pyindex

[ERROR] in line 126: exit code 0: while executing command $STD pip3 install -r /opt/frigate/docker/main/requirements.txt
/dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context

@hmronline commented on GitHub (Dec 2, 2025): **Tested on**: Hardware: Custom PC, Intel i7-10700F, 64 GB RAM GPU: NVIDIA GeForce GTX 1650 TPU: None Host: Proxmox 9.1.1 Container: Proxmox LXC (Debian 12, unprivileged) **Script**: `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"` **Config**: ``` root@pve:~# cat /opt/community-scripts/frigate.conf # frigate Configuration File # Generated on Tue Dec 2 06:42:12 PM -03 2025 CT_TYPE="1" DISK_SIZE="50" CORE_COUNT="4" RAM_SIZE="4096" VERBOSE="yes" PW="secret" #CT_ID=100 HN="frigate" BRG="vmbr0" NET="dhcp" IPV6_METHOD="none" # Set this only if using "IPV6_METHOD=static" #IPV6STATIC="fd00::1234/64" GATE="none" APT_CACHER_IP="none" MTU="1500" SD="none" NS="none" MAC="none" VLAN="10" SSH="no" SSH_AUTHORIZED_KEY="" TAGS="remz1337;nvr" ENABLE_FUSE="no" ENABLE_TUN="no" ``` **Error** (just the error part): ``` Building wheels for collected packages: nvidia-pyindex Building wheel for nvidia-pyindex (pyproject.toml) ... error error: subprocess-exited-with-error × Building wheel for nvidia-pyindex (pyproject.toml) did not run successfully. │ exit code: 1 ╰─> [99 lines of output] /tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/dist.py:759: SetuptoolsDeprecationWarning: License classifiers are deprecated. !! ******************************************************************************** Please consider removing the following classifiers in favor of a SPDX license expression: License :: OSI Approved :: Apache Software License See https://packaging.python.org/en/latest/guides/writing-pyproject-toml/#license for details. ******************************************************************************** !! self._finalize_license_expression() running bdist_wheel running build running build_py creating build/lib/nvidia_pyindex copying nvidia_pyindex/cmdline.py -> build/lib/nvidia_pyindex copying nvidia_pyindex/__init__.py -> build/lib/nvidia_pyindex copying nvidia_pyindex/utils.py -> build/lib/nvidia_pyindex running egg_info writing nvidia_pyindex.egg-info/PKG-INFO writing dependency_links to nvidia_pyindex.egg-info/dependency_links.txt writing entry points to nvidia_pyindex.egg-info/entry_points.txt writing top-level names to nvidia_pyindex.egg-info/top_level.txt reading manifest file 'nvidia_pyindex.egg-info/SOURCES.txt' reading manifest template 'MANIFEST.in' adding license file 'LICENSE.md' writing manifest file 'nvidia_pyindex.egg-info/SOURCES.txt' installing to build/bdist.linux-x86_64/wheel running install %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% COMMAND: InstallCommand %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Traceback (most recent call last): File "/tmp/pip-install-984lbo3f/nvidia-pyindex_3d7ae6811f234c6fbdeb80137ea7ec7c/nvidia_pyindex/utils.py", line 20, in get_configuration_files output = subprocess.check_output(command, stderr=subprocess.STDOUT) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/lib/python3.11/subprocess.py", line 466, in check_output return run(*popenargs, stdout=PIPE, timeout=timeout, check=True, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/lib/python3.11/subprocess.py", line 571, in run raise CalledProcessError(retcode, process.args, subprocess.CalledProcessError: Command '['/usr/bin/python3', '-c', 'from pip._internal.configuration import get_configuration_files; print(get_configuration_files())']' returned non-zero exit status 1. The above exception was the direct cause of the following exception: Traceback (most recent call last): File "/usr/local/lib/python3.11/dist-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 389, in <module> main() File "/usr/local/lib/python3.11/dist-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 373, in main json_out["return_val"] = hook(**hook_input["kwargs"]) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.11/dist-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 280, in build_wheel return _build_backend().build_wheel( ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/build_meta.py", line 435, in build_wheel return _build(['bdist_wheel', '--dist-info-dir', str(metadata_directory)]) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/build_meta.py", line 423, in _build return self._build_with_temp_dir( ^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/build_meta.py", line 404, in _build_with_temp_dir self.run_setup() File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/build_meta.py", line 512, in run_setup super().run_setup(setup_script=setup_script) File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/build_meta.py", line 317, in run_setup exec(code, locals()) File "<string>", line 238, in <module> File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/__init__.py", line 115, in setup return distutils.core.setup(**attrs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/_distutils/core.py", line 186, in setup return run_commands(dist) ^^^^^^^^^^^^^^^^^^ File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/_distutils/core.py", line 202, in run_commands dist.run_commands() File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/_distutils/dist.py", line 1002, in run_commands self.run_command(cmd) File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/dist.py", line 1102, in run_command super().run_command(command) File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/_distutils/dist.py", line 1021, in run_command cmd_obj.run() File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/command/bdist_wheel.py", line 405, in run self.run_command("install") File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/_distutils/cmd.py", line 357, in run_command self.distribution.run_command(command) File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/dist.py", line 1102, in run_command super().run_command(command) File "/tmp/pip-build-env-1vh6wjkj/overlay/local/lib/python3.11/dist-packages/setuptools/_distutils/dist.py", line 1021, in run_command cmd_obj.run() File "<string>", line 172, in run File "<string>", line 71, in _install_nvidia_pypi_index File "/tmp/pip-install-984lbo3f/nvidia-pyindex_3d7ae6811f234c6fbdeb80137ea7ec7c/nvidia_pyindex/utils.py", line 24, in get_configuration_files raise RuntimeError(output) from e RuntimeError: Traceback (most recent call last): File "<string>", line 1, in <module> ModuleNotFoundError: No module named 'pip' [end of output] note: This error originates from a subprocess, and is likely not a problem with pip. ERROR: Failed building wheel for nvidia-pyindex Failed to build nvidia-pyindex error: failed-wheel-build-for-install × Failed to build installable wheels for some pyproject.toml based projects ╰─> nvidia-pyindex [ERROR] in line 126: exit code 0: while executing command $STD pip3 install -r /opt/frigate/docker/main/requirements.txt /dev/fd/63: line 36: pop_var_context: head of shell_variables not a function context ```

kerem commented

Author

Owner

@remz1337 commented on GitHub (Dec 3, 2025):

I've pushed a fix to remove the problematic nvidia-pyindex line. Tested with priviledged and no gpu passthrough and it successfully installed everything.

bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"

Not sure why you are all facing issues, but clearly there's a difference in our setups. I'm still on PVE 8.4, so might be that... (yea, I'll update soon...)

@remz1337 commented on GitHub (Dec 3, 2025): I've pushed a fix to remove the problematic `nvidia-pyindex` line. Tested with priviledged and no gpu passthrough and it successfully installed everything. `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"` Not sure why you are all facing issues, but clearly there's a difference in our setups. I'm still on PVE 8.4, so might be that... (yea, I'll update soon...)

kerem commented

Author

Owner

@Honusnap commented on GitHub (Dec 3, 2025):

I've pushed a fix to remove the problematic nvidia-pyindex line. Tested with priviledged and no gpu passthrough and it successfully installed everything.

bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"

Not sure why you are all facing issues, but clearly there's a difference in our setups. I'm still on PVE 8.4, so might be that... (yea, I'll update soon...)

Off topic but are you going with the update script .. or backup install the 9 fresh and restore ?
I'm in the same boat .. and .. i'm lazy, i think i will backup, try the update by script .. then if it fail ... go the fresh road.

@Honusnap commented on GitHub (Dec 3, 2025): > I've pushed a fix to remove the problematic `nvidia-pyindex` line. Tested with priviledged and no gpu passthrough and it successfully installed everything. > > `bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)"` > > Not sure why you are all facing issues, but clearly there's a difference in our setups. I'm still on PVE 8.4, so might be that... (yea, I'll update soon...) Off topic but are you going with the update script .. or backup install the 9 fresh and restore ? I'm in the same boat .. and .. i'm lazy, i think i will backup, try the update by script .. then if it fail ... go the fresh road.

kerem commented

Author

Owner

@Pehun commented on GitHub (Dec 4, 2025):

Proxmox 8.2, intel AR380,
when I change the detector from AUTO to GPU, i get :
2025-12-04 12:02:17.766191 [INFO] Preparing Frigate...
2025-12-04 12:02:18.113809 [INFO] Starting Frigate...
2025-12-04 12:02:20.739177 [2025-12-04 12:02:20] frigate.util.config INFO : Checking if frigate config needs migration...
2025-12-04 12:02:20.745749 [2025-12-04 12:02:20] frigate.util.config INFO : frigate config does not need migration...
2025-12-04 12:02:20.758970 [2025-12-04 12:02:20] frigate.app INFO : Starting Frigate (0.16.2-)
2025-12-04 12:02:20.764775 [2025-12-04 12:02:20] peewee_migrate.logs INFO : Starting migrations
2025-12-04 12:02:20.765763 [2025-12-04 12:02:20] peewee_migrate.logs INFO : There is nothing to migrate
2025-12-04 12:02:20.776972 [2025-12-04 12:02:20] frigate.app INFO : Recording process started: 47110
2025-12-04 12:02:20.777013 [2025-12-04 12:02:20] frigate.app INFO : Review process started: 47113
2025-12-04 12:02:20.779233 [2025-12-04 12:02:20] frigate.app INFO : go2rtc process pid: 46351
2025-12-04 12:02:20.791469 [2025-12-04 12:02:20] detector.detector01 INFO : Starting detection process: 47135
2025-12-04 12:02:20.799281 [2025-12-04 12:02:20] frigate.app INFO : Output process started: 47151
2025-12-04 12:02:20.811004 [2025-12-04 12:02:20] frigate.app INFO : Camera processor started for test: 47175
2025-12-04 12:02:20.825912 [2025-12-04 12:02:20] frigate.app INFO : Capture process started for test: 47185
2025-12-04 12:02:20.868763 [2025-12-04 12:02:20] frigate.video ERROR : test: Unable to read frames from ffmpeg process.
2025-12-04 12:02:20.869005 [2025-12-04 12:02:20] frigate.video ERROR : test: ffmpeg process is not running. exiting capture thread...
2025-12-04 12:02:20.929420 [2025-12-04 12:02:20] frigate.api.fastapi_app INFO : Starting FastAPI app
2025-12-04 12:02:21.104753 [2025-12-04 12:02:21] frigate.api.fastapi_app INFO : FastAPI started
2025-12-04 12:02:40.838153 [2025-12-04 12:02:40] watchdog.test ERROR : Ffmpeg process crashed unexpectedly for test.
2025-12-04 12:02:40.838381 [2025-12-04 12:02:40] watchdog.test ERROR : The following ffmpeg logs include the last 100 lines prior to exit.
2025-12-04 12:02:40.838574 [2025-12-04 12:02:40] ffmpeg.test.detect ERROR : [AVFormatContext @ 0x59d101434a80] Unable to choose an output format for 'GPU'; use a standard extension for the filename or specify the format manually.
2025-12-04 12:02:40.838746 [2025-12-04 12:02:40] ffmpeg.test.detect ERROR : [out#0 @ 0x59d101273840] Error initializing the muxer for GPU: Invalid argument
2025-12-04 12:02:40.838898 [2025-12-04 12:02:40] ffmpeg.test.detect ERROR : Error opening output file GPU.
2025-12-04 12:02:40.839083 [2025-12-04 12:02:40] ffmpeg.test.detect ERROR : Error opening output files: Invalid argument
2025-12-04 12:02:40.839234 [2025-12-04 12:02:40] watchdog.test INFO : Restarting ffmpeg...
2025-12-04 12:02:40.853947 [2025-12-04 12:02:40] frigate.video ERROR : test: Unable to read frames from ffmpeg process.
2025-12-04 12:02:40.854330 [2025-12-04 12:02:40] frigate.video ERROR : test: ffmpeg process is not running. exiting capture thread...

@Pehun commented on GitHub (Dec 4, 2025): Proxmox 8.2, intel AR380, when I change the detector from AUTO to GPU, i get : 2025-12-04 12:02:17.766191 [INFO] Preparing Frigate... 2025-12-04 12:02:18.113809 [INFO] Starting Frigate... 2025-12-04 12:02:20.739177 [2025-12-04 12:02:20] frigate.util.config INFO : Checking if frigate config needs migration... 2025-12-04 12:02:20.745749 [2025-12-04 12:02:20] frigate.util.config INFO : frigate config does not need migration... 2025-12-04 12:02:20.758970 [2025-12-04 12:02:20] frigate.app INFO : Starting Frigate (0.16.2-) 2025-12-04 12:02:20.764775 [2025-12-04 12:02:20] peewee_migrate.logs INFO : Starting migrations 2025-12-04 12:02:20.765763 [2025-12-04 12:02:20] peewee_migrate.logs INFO : There is nothing to migrate 2025-12-04 12:02:20.776972 [2025-12-04 12:02:20] frigate.app INFO : Recording process started: 47110 2025-12-04 12:02:20.777013 [2025-12-04 12:02:20] frigate.app INFO : Review process started: 47113 2025-12-04 12:02:20.779233 [2025-12-04 12:02:20] frigate.app INFO : go2rtc process pid: 46351 2025-12-04 12:02:20.791469 [2025-12-04 12:02:20] detector.detector01 INFO : Starting detection process: 47135 2025-12-04 12:02:20.799281 [2025-12-04 12:02:20] frigate.app INFO : Output process started: 47151 2025-12-04 12:02:20.811004 [2025-12-04 12:02:20] frigate.app INFO : Camera processor started for test: 47175 2025-12-04 12:02:20.825912 [2025-12-04 12:02:20] frigate.app INFO : Capture process started for test: 47185 2025-12-04 12:02:20.868763 [2025-12-04 12:02:20] frigate.video ERROR : test: Unable to read frames from ffmpeg process. 2025-12-04 12:02:20.869005 [2025-12-04 12:02:20] frigate.video ERROR : test: ffmpeg process is not running. exiting capture thread... 2025-12-04 12:02:20.929420 [2025-12-04 12:02:20] frigate.api.fastapi_app INFO : Starting FastAPI app 2025-12-04 12:02:21.104753 [2025-12-04 12:02:21] frigate.api.fastapi_app INFO : FastAPI started 2025-12-04 12:02:40.838153 [2025-12-04 12:02:40] watchdog.test ERROR : Ffmpeg process crashed unexpectedly for test. 2025-12-04 12:02:40.838381 [2025-12-04 12:02:40] watchdog.test ERROR : The following ffmpeg logs include the last 100 lines prior to exit. 2025-12-04 12:02:40.838574 [2025-12-04 12:02:40] ffmpeg.test.detect ERROR : [AVFormatContext @ 0x59d101434a80] Unable to choose an output format for 'GPU'; use a standard extension for the filename or specify the format manually. 2025-12-04 12:02:40.838746 [2025-12-04 12:02:40] ffmpeg.test.detect ERROR : [out#0 @ 0x59d101273840] Error initializing the muxer for GPU: Invalid argument 2025-12-04 12:02:40.838898 [2025-12-04 12:02:40] ffmpeg.test.detect ERROR : Error opening output file GPU. 2025-12-04 12:02:40.839083 [2025-12-04 12:02:40] ffmpeg.test.detect ERROR : Error opening output files: Invalid argument 2025-12-04 12:02:40.839234 [2025-12-04 12:02:40] watchdog.test INFO : Restarting ffmpeg... 2025-12-04 12:02:40.853947 [2025-12-04 12:02:40] frigate.video ERROR : test: Unable to read frames from ffmpeg process. 2025-12-04 12:02:40.854330 [2025-12-04 12:02:40] frigate.video ERROR : test: ffmpeg process is not running. exiting capture thread...

kerem commented

Author

Owner

@YugiFanGX commented on GitHub (Dec 4, 2025):

I can't get face recognition to work. When I am activating it under settings, I can't load the Web UI anymore. Is someone else have the same issue? are there some unmet dependencies?

@YugiFanGX commented on GitHub (Dec 4, 2025): I can't get face recognition to work. When I am activating it under settings, I can't load the Web UI anymore. Is someone else have the same issue? are there some unmet dependencies?

kerem commented

Author

Owner

@fjleon1980 commented on GitHub (Dec 4, 2025):

the latest version of bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)" did work for me and frigate is running, however performance is really bad with hwaccel auto. I also tried preset-vaapi (maybe the same as auto) with the same performance. The preset with intel quickstart throws a ffmpeg error. Anyone has experienced this on Intel N150?

@fjleon1980 commented on GitHub (Dec 4, 2025): the latest version of bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/remz/ct/frigate.sh)" did work for me and frigate is running, however performance is really bad with hwaccel auto. I also tried preset-vaapi (maybe the same as auto) with the same performance. The preset with intel quickstart throws a ffmpeg error. Anyone has experienced this on Intel N150?

kerem commented

Author

Owner

@remz1337 commented on GitHub (Dec 5, 2025):

@fjleon1980 did you run with default config? for iGPU to work you need to use advanced settings: priviledged + disable nvidia passthrough

@remz1337 commented on GitHub (Dec 5, 2025): @fjleon1980 did you run with default config? for iGPU to work you need to use advanced settings: priviledged + disable nvidia passthrough

kerem commented

Author

Owner

@fjleon1980 commented on GitHub (Dec 5, 2025):

@fjleon1980 did you run with default config? for iGPU to work you need to use advanced settings: priviledged + disable nvidia passthrough

yes, did that, then opened the parking video, which makes it big, then enabled detection. cpu goes up to nearly 60%. i found interesting that hwaccel is set to auto, which i couldn't find in the documentation, which actually recommends to use preset-intel-qsv-h264 but if i put that ffmpeg crashes. either auto or preset-vaapi work but it's too slow despite giving it 4 cores and 4 gigs of ram with no cameras other than the test parking video.

I think i am going to try with docker and full blown qemu VM in proxmox to compare

@fjleon1980 commented on GitHub (Dec 5, 2025): > [@fjleon1980](https://github.com/fjleon1980) did you run with default config? for iGPU to work you need to use advanced settings: priviledged + disable nvidia passthrough yes, did that, then opened the parking video, which makes it big, then enabled detection. cpu goes up to nearly 60%. i found interesting that hwaccel is set to auto, which i couldn't find in the documentation, which actually recommends to use preset-intel-qsv-h264 but if i put that ffmpeg crashes. either auto or preset-vaapi work but it's too slow despite giving it 4 cores and 4 gigs of ram with no cameras other than the test parking video. I think i am going to try with docker and full blown qemu VM in proxmox to compare

kerem commented

Author

Owner

@Zixim commented on GitHub (Dec 5, 2025):

@fjleon1980 I, also, ran into issues with face recognition. I finally went with the comminuty-scripts docker LXC + official Frigate docker container.
Best of both worlds and the thing runs stable as a train, with iGPU & Coral TPU.

@Zixim commented on GitHub (Dec 5, 2025): @fjleon1980 I, also, ran into issues with face recognition. I finally went with the comminuty-scripts docker LXC + official Frigate docker container. Best of both worlds and the thing runs stable as a train, with iGPU & Coral TPU.

kerem commented

Author

Owner

@barryvdh commented on GitHub (Dec 5, 2025):

@fjleon1980 I, also, ran into issues with face recognition. I finally went with the comminuty-scripts docker LXC + official Frigate docker container. Best of both worlds and the thing runs stable as a train, with iGPU & Coral TPU.

Did you do priviliged or unprivileged? I'm using an N150 so will also try this route.

@barryvdh commented on GitHub (Dec 5, 2025): > [@fjleon1980](https://github.com/fjleon1980) I, also, ran into issues with face recognition. I finally went with the comminuty-scripts docker LXC + official Frigate docker container. Best of both worlds and the thing runs stable as a train, with iGPU & Coral TPU. Did you do priviliged or unprivileged? I'm using an N150 so will also try this route.

kerem commented

Author

Owner

@barryvdh commented on GitHub (Dec 5, 2025):

I did the following for Beelink EQ14 with N150

Make sure Proxmox is recent (Proxmox 9, recent Debian 13)
Use bios settings from https://github.com/patcfly/n150-passthrough
Install Intel drivers on Proxmox host from: https://github.com/intel/compute-runtime/releases
Install Docker LXC from the community scripts, run as privileged. This detected the Intel driver and configured passthrough
Upgraded the LXC from Debian 12 to 13
Install Docker
Install Frigate for Docker
Enable OpenVino + GPU

Now Frigate detects the Intel GPU and detector interference is 10ms (1 camera, 300x300)

@barryvdh commented on GitHub (Dec 5, 2025): I did the following for Beelink EQ14 with N150 - Make sure Proxmox is recent (Proxmox 9, recent Debian 13) - Use bios settings from https://github.com/patcfly/n150-passthrough - Install Intel drivers on Proxmox host from: https://github.com/intel/compute-runtime/releases - Install Docker LXC from the community scripts, run as privileged. This detected the Intel driver and configured passthrough - Upgraded the LXC from Debian 12 to 13 - Install Docker - Install Frigate for Docker - Enable OpenVino + GPU Now Frigate detects the Intel GPU and detector interference is 10ms (1 camera, 300x300)

kerem commented

Author

Owner

@fjleon1980 commented on GitHub (Dec 5, 2025):

I did the following for Beelink EQ14 with N150

* Make sure Proxmox is recent (Proxmox 9, recent Debian 13)

* Use bios settings from https://github.com/patcfly/n150-passthrough

* Install Intel drivers on Proxmox host from: https://github.com/intel/compute-runtime/releases

* Install Docker LXC from the community scripts, run as privileged. This detected the Intel driver and configured passthrough

* Upgraded the LXC from Debian 12 to 13

* Install Docker

* Install Frigate for Docker

* Enable OpenVino + GPU

Now Frigate detects the Intel GPU and detector interference is 10ms (1 camera, 300x300)

I tried the docker lxc method today, which gives you a blank frigate file (not even the parking video). I copied the video from the LXC created from this frigate community script. Honestly, i don't see any difference. Above you say install docker twice, which I didn't understand. I just installed the docker LXC, chose to use portainer, then copy pasted the docker compose command from the frigate docs.

I see intel video being reported on both containers. My issue is that I can't seem to be able to use the intel quicksync preset (recommended for N150) as ffmpeg just says invalid argument and instead have to use preset-vaapi and i wanted to see how it performed

The steps on the patcfly docs is for QEMU VM, which is the official proxmox recommendation for containers

Here are my stats after a few seconds of running detection on the parking video

I was hoping to see less CPU than that. not sure why igpu is overall low. the inference speed is fine at 10 ms

@fjleon1980 commented on GitHub (Dec 5, 2025): > I did the following for Beelink EQ14 with N150 > > * Make sure Proxmox is recent (Proxmox 9, recent Debian 13) > > * Use bios settings from https://github.com/patcfly/n150-passthrough > > * Install Intel drivers on Proxmox host from: https://github.com/intel/compute-runtime/releases > > * Install Docker LXC from the community scripts, run as privileged. This detected the Intel driver and configured passthrough > > * Upgraded the LXC from Debian 12 to 13 > > * Install Docker > > * Install Frigate for Docker > > * Enable OpenVino + GPU > > > Now Frigate detects the Intel GPU and detector interference is 10ms (1 camera, 300x300) I tried the docker lxc method today, which gives you a blank frigate file (not even the parking video). I copied the video from the LXC created from this frigate community script. Honestly, i don't see any difference. Above you say install docker twice, which I didn't understand. I just installed the docker LXC, chose to use portainer, then copy pasted the docker compose command from the frigate docs. I see intel video being reported on both containers. My issue is that I can't seem to be able to use the intel quicksync preset (recommended for N150) as ffmpeg just says invalid argument and instead have to use preset-vaapi and i wanted to see how it performed The steps on the patcfly docs is for QEMU VM, which is the official proxmox recommendation for containers Here are my stats after a few seconds of running detection on the parking video <img width="1208" height="651" alt="Image" src="https://github.com/user-attachments/assets/d21176be-addc-4a02-9e4b-819f569ec6cf" /> I was hoping to see less CPU than that. not sure why igpu is overall low. the inference speed is fine at 10 ms

kerem commented

Author

Owner

@barryvdh commented on GitHub (Dec 6, 2025):

I tried the docker lxc method today, which gives you a blank frigate file (not even the parking video). I copied the video from the LXC created from this frigate community script. Honestly, i don't see any difference. Above you say install docker twice, which I didn't understand. I just installed the docker LXC, chose to use portainer, then copy pasted the docker compose command from the frigate docs.

I see intel video being reported on both containers. My issue is that I can't seem to be able to use the intel quicksync preset (recommended for N150) as ffmpeg just says invalid argument and instead have to use preset-vaapi and i wanted to see how it performed

The steps on the patcfly docs is for QEMU VM, which is the official proxmox recommendation for containers

Here are my stats after a few seconds of running detection on the parking video
I was hoping to see less CPU than that. not sure why igpu is overall low. the inference speed is fine at 10 ms

sorry, yes not sure why I did that. Was trying a few things before it worked. But just updated the Docker container to Debian 13, not sure it that was required though.

Fpr the patchfly repo, I only used the BIOS settings from there, had to enable SR-IOV Support and/or set Internal Graphics to Enabled instead of auto + set DVMT to 64MB. The Intel VT-d was already enabled.

I'm not sure about QuickSync, but preset-vaapi is working for me it seems.

@barryvdh commented on GitHub (Dec 6, 2025): > > I tried the docker lxc method today, which gives you a blank frigate file (not even the parking video). I copied the video from the LXC created from this frigate community script. Honestly, i don't see any difference. Above you say install docker twice, which I didn't understand. I just installed the docker LXC, chose to use portainer, then copy pasted the docker compose command from the frigate docs. > > I see intel video being reported on both containers. My issue is that I can't seem to be able to use the intel quicksync preset (recommended for N150) as ffmpeg just says invalid argument and instead have to use preset-vaapi and i wanted to see how it performed > > The steps on the patcfly docs is for QEMU VM, which is the official proxmox recommendation for containers > > Here are my stats after a few seconds of running detection on the parking video > > <img alt="Image" width="1208" height="651" src="https://private-user-images.githubusercontent.com/23721392/523195854-d21176be-addc-4a02-9e4b-819f569ec6cf.png?jwt=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.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.dcHWFN8rV8D3z8sNnlaui2_JvL2efdmoVf-qkym0iqg"> > I was hoping to see less CPU than that. not sure why igpu is overall low. the inference speed is fine at 10 ms sorry, yes not sure why I did that. Was trying a few things before it worked. But just updated the Docker container to Debian 13, not sure it that was required though. Fpr the patchfly repo, I only used the BIOS settings from there, had to enable SR-IOV Support and/or set Internal Graphics to Enabled instead of auto + set DVMT to 64MB. The Intel VT-d was already enabled. I'm not sure about QuickSync, but preset-vaapi is working for me it seems.

kerem commented

Author

Owner

@fjleon1980 commented on GitHub (Dec 6, 2025):

one thing that i found odd with this script is that I do not see a device passthrough on the LXC container (despite using the advanced setup):

If i deploy the docker LXC community script, I do:

i wonder if this needs to be changed

@fjleon1980 commented on GitHub (Dec 6, 2025): one thing that i found odd with this script is that I do not see a device passthrough on the LXC container (despite using the advanced setup): <img width="523" height="146" alt="Image" src="https://github.com/user-attachments/assets/e5cb7c4f-fd58-476f-a686-0ccde3d24d24" /> If i deploy the docker LXC community script, I do: <img width="474" height="76" alt="Image" src="https://github.com/user-attachments/assets/e8bedb0f-0d6b-4d98-a5ef-bba1e19de522" /> i wonder if this needs to be changed

kerem commented

Author

Owner

@Pehun commented on GitHub (Dec 10, 2025):

I tried what said above,
Use the helper script tu create docker lxc
it even propose to pass GPU!
And it worked better than ever!

@Pehun commented on GitHub (Dec 10, 2025): I tried what said above, Use the helper script tu create docker lxc it even propose to pass GPU! And it worked better than ever!

kerem commented

Author

Owner

@Patrick010 commented on GitHub (Dec 10, 2025):

Isnt the goal of these scripts to not to use Docker?

@Patrick010 commented on GitHub (Dec 10, 2025): Isnt the goal of these scripts to _not_ to use Docker?

kerem commented

Author

Owner

@Pehun commented on GitHub (Dec 10, 2025):

If i'm not wrong it uses docker anyway.
the only step adding with the frigate script is the addition of the frigate docker

@Pehun commented on GitHub (Dec 10, 2025): If i'm not wrong it uses docker anyway. the only step adding with the frigate script is the addition of the frigate docker

kerem commented

Author

Owner

@Pehun commented on GitHub (Dec 10, 2025):

to condense :
use this script :
bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/ct/docker.sh)"
choose advanced

Priviledged
4096 ram
20GO storage

In the LXC config, verify renderd128 is passed and add your coral :
lxc.mount.entry: /dev/bus/usb/002/ dev/bus/usb/002/ none bind,optional,create=dir 0,0

In the LXC
Create /opt/frigate/config/
do nano compose docker-compose.yaml in /opt/

paste this (adapt to your config)

a mettre dans repertoire /opt

lancer par "docker-compose up -d"

version: '3.9'

services:

frigate:
container_name: frigate
privileged: true
restart: unless-stopped
image: ghcr.io/blakeblackshear/frigate:stable
shm_size: "256mb" # update for your cameras based on calculation above
devices:
- /dev/bus/usb:/dev/bus/usb
- /dev/dri/renderD128 # for intel hwaccel, needs to be updated for your hardware
volumes:
- /etc/localtime:/etc/localtime:ro
- /opt/frigate/config:/config
- /mnt/frigate:/media/frigate
- type: tmpfs # Optional: 1GB of memory, reduces SSD/SD Card wear
target: /tmp/cache
tmpfs:
size: 1000000000
ports:
- "5000:5000"
- "1935:1935" # RTMP feeds

For coral, execute this in the lxc :
echo "deb https://packages.cloud.google.com/apt coral-edgetpu-stable main" | sudo tee /etc/apt/sources.list.d/coral-edgetpu.list
curl https://packages.cloud.google.com/apt/doc/apt-key.gpg | sudo apt-key add -
sudo apt-get update
sudo apt-get install libedgetpu1-std

Once all done, do this in /opt
docker compose -f docker-compose.yaml up -d
and it's finished

@Pehun commented on GitHub (Dec 10, 2025): to condense : use this script : bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/ct/docker.sh)" choose advanced - Priviledged - 4096 ram - 20GO storage In the LXC config, verify renderd128 is passed and add your coral : lxc.mount.entry: /dev/bus/usb/002/ dev/bus/usb/002/ none bind,optional,create=dir 0,0 In the LXC Create /opt/frigate/config/ do nano compose docker-compose.yaml in /opt/ paste this (adapt to your config) # a mettre dans repertoire /opt # lancer par "docker-compose up -d" version: '3.9' services: frigate: container_name: frigate privileged: true restart: unless-stopped image: ghcr.io/blakeblackshear/frigate:stable shm_size: "256mb" # update for your cameras based on calculation above devices: - /dev/bus/usb:/dev/bus/usb - /dev/dri/renderD128 # for intel hwaccel, needs to be updated for your hardware volumes: - /etc/localtime:/etc/localtime:ro - /opt/frigate/config:/config - /mnt/frigate:/media/frigate - type: tmpfs # Optional: 1GB of memory, reduces SSD/SD Card wear target: /tmp/cache tmpfs: size: 1000000000 ports: - "5000:5000" - "1935:1935" # RTMP feeds For coral, execute this in the lxc : echo "deb https://packages.cloud.google.com/apt coral-edgetpu-stable main" | sudo tee /etc/apt/sources.list.d/coral-edgetpu.list curl https://packages.cloud.google.com/apt/doc/apt-key.gpg | sudo apt-key add - sudo apt-get update sudo apt-get install libedgetpu1-std Once all done, do this in /opt docker compose -f docker-compose.yaml up -d and it's finished

kerem commented

Author

Owner

@MarcSeebold commented on GitHub (Dec 10, 2025):

Be careful, this community script (https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/ct/docker.sh) does "apt-get -y upgrade" which IIRC is not recommended by Proxmox. Make backups!

Also, the recommended way to run Docker on Proxmox is in a VM. This provides proper isolation at the cost of mounting the GPU. One overly complicated and mediocre solution is vGPU. Or, if you can afford it, do a PCIe passthrough of your GPU. Then, it's exclusively used by that VM.

@MarcSeebold commented on GitHub (Dec 10, 2025): Be careful, this community script (https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/ct/docker.sh) does "apt-get -y upgrade" which IIRC is not recommended by Proxmox. Make backups! Also, the recommended way to run Docker on Proxmox is in a VM. This provides proper isolation at the cost of mounting the GPU. One overly complicated and mediocre solution is vGPU. Or, if you can afford it, do a PCIe passthrough of your GPU. Then, it's exclusively used by that VM.

kerem commented

Author

Owner

@MarcSeebold commented on GitHub (Dec 10, 2025):

Has anyone tried converting the Frigate Docker to LXC? https://github.com/glennswest/container2lxc

@MarcSeebold commented on GitHub (Dec 10, 2025): Has anyone tried converting the Frigate Docker to LXC? https://github.com/glennswest/container2lxc

kerem commented

Author

Owner

@Zixim commented on GitHub (Dec 11, 2025):

Be careful, this community script (https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/ct/docker.sh) does "apt-get -y upgrade" which IIRC is not recommended by Proxmox. Make backups!

No, it does not run apt-get -y upgrade on the Proxmox host. apt-get -y upgrade is part of the update function which you can run in the LXC via the command update if/when you want to update the LXC.

@Zixim commented on GitHub (Dec 11, 2025): > Be careful, this community script (https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/ct/docker.sh) does "apt-get -y upgrade" which IIRC is not recommended by Proxmox. Make backups! > No, it does not run `apt-get -y upgrade` on the Proxmox host. `apt-get -y upgrade` is part of the **update function** which you can run in the LXC via the command `update` if/when you want to update the LXC.

kerem commented

Author

Owner

@J-Smits commented on GitHub (Jan 5, 2026):

Good morning,

I'm using now the test script from Remz1337, a month now.

bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/main/ct/frigate.sh)"

I'm running it on a HPE microserver gen10 with the Xeon processor under Proxmox 9.1.4 and it is working correct now, was able to get a running Frigate instance, working with the USB coral, i only noticed that the audio isn't working anymore with this script, if i check the log i see the following message:

are there maybe more people using this script right now and they have the audio detector working?

@J-Smits commented on GitHub (Jan 5, 2026): Good morning, I'm using now the test script from Remz1337, a month now. bash -c "$(wget -qLO - https://github.com/remz1337/ProxmoxVE/raw/main/ct/frigate.sh)" I'm running it on a HPE microserver gen10 with the Xeon processor under Proxmox 9.1.4 and it is working correct now, was able to get a running Frigate instance, working with the USB coral, i only noticed that the audio isn't working anymore with this script, if i check the log i see the following message: unknown | 2025-12-31 13:45:27 | unknown | File "/opt/frigate/frigate/events/audio.py", line 131, in __init__ unknown | 2025-12-31 13:45:27 | unknown | self.detector = AudioTfl(stop_event, self.config.audio.num_threads) are there maybe more people using this script right now and they have the audio detector working?

kerem commented

Author

Owner

@croessi commented on GitHub (Jan 10, 2026):

Hi,

I created a script based on the previous LXC Frigate Scripts. I was able to update all the reference to the lates mentioned in the dockerfile and have successfully tested this for installing (frigate 16.0) and updating to 16.3 in a fresh lxc container. I am succesfully using this with a coral USB accelerator. Maybe this helps others.

install_frigate.sh

@croessi commented on GitHub (Jan 10, 2026): Hi, I created a script based on the previous LXC Frigate Scripts. I was able to update all the reference to the lates mentioned in the dockerfile and have successfully tested this for installing (frigate 16.0) and updating to 16.3 in a fresh lxc container. I am succesfully using this with a coral USB accelerator. Maybe this helps others. [install_frigate.sh](https://github.com/user-attachments/files/24540598/install_frigate.sh)

kerem commented

Author

Owner

@Patrick010 commented on GitHub (Jan 12, 2026):

Guys,

I've struggled myself through the conversations in this topic and tried to install Frigate in a Proxmox 9.0.14 LXC, but none seem to work. Which is the proper script to use? People seem to have different experiences and results, but I believe there is no near production ready version available yet?
The last one I ran was https://github.com/community-scripts/ProxmoxVED/pull/929, but that tries to install on Debian 11?

@Patrick010 commented on GitHub (Jan 12, 2026): Guys, I've struggled myself through the conversations in this topic and tried to install Frigate in a Proxmox 9.0.14 LXC, but none seem to work. Which is the proper script to use? People seem to have different experiences and results, but I believe there is no near production ready version available yet? The last one I ran was https://github.com/community-scripts/ProxmoxVED/pull/929, but that tries to install on Debian 11?

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Jan 12, 2026):

I tried for months to get the script working well for me. Provided some feedback at times. But I always end up with a system that just didn't work well. Frigate would crash, etc. I then saw the message from Pehen on December 10th and tried that. That seemed to work, then I made some changes to my frigate configuration after that. Been working fine. Not sure if it was the config changes or the different install process but it's working and I only made the config changes after running for a couple days on a stable system.

@atrozgrifo commented on GitHub (Jan 12, 2026): I tried for months to get the script working well for me. Provided some feedback at times. But I always end up with a system that just didn't work well. Frigate would crash, etc. I then saw the message from Pehen on December 10th and tried that. That seemed to work, then I made some changes to my frigate configuration after that. Been working fine. Not sure if it was the config changes or the different install process but it's working and I only made the config changes after running for a couple days on a stable system.

kerem commented

Author

Owner

@rustikles commented on GitHub (Jan 12, 2026):

In the interest of having something that works, maybe the best option is pivot to a VM setup script. It's the recommended method by frigate developers and eliminates a lot of these dependency/LXC intricacies. That was the path I took when it came time to upgrade from an ancient LXC frigate to something current.

@rustikles commented on GitHub (Jan 12, 2026): In the interest of having _something_ that works, maybe the best option is pivot to a VM setup script. It's the recommended method by frigate developers and eliminates a lot of these dependency/LXC intricacies. That was the path I took when it came time to upgrade from an ancient LXC frigate to something current.

kerem commented

Author

Owner

@albotroz commented on GitHub (Jan 12, 2026):

Why was version 0.14 removed from the community scripts mainpage? The version still works for me—without any problems.

Were there any serious issues that led to it being removed?

Is there no answer to this?
I'm still running 0.14.

Why wasn't 0.14 left in place until 0.15 or 0.16 was made bulletproof for the Proxmox environment?

I don't understand this approach.
Apparently, Frigate 0.15/0.16 is causing problems for many people within a Proxmox environment.
Frigate 0.15 or 0.16 is probably one of the services that causes the most problems... from what I can see.

Of course, I would like to upgrade to 0.15 or 0.16, but what I'm reading here is putting me off.

@albotroz commented on GitHub (Jan 12, 2026): > Why was version 0.14 removed from the community scripts mainpage? The version still works for me—without any problems. > > Were there any serious issues that led to it being removed? Is there no answer to this? I'm still running 0.14. Why wasn't 0.14 left in place until 0.15 or 0.16 was made bulletproof for the Proxmox environment? I don't understand this approach. Apparently, Frigate 0.15/0.16 is causing problems for many people within a Proxmox environment. Frigate 0.15 or 0.16 is probably one of the services that causes the most problems... from what I can see. Of course, I would like to upgrade to 0.15 or 0.16, but what I'm reading here is putting me off.

kerem commented

Author

Owner

@atrozgrifo commented on GitHub (Jan 12, 2026):

I was running Frigate 0.16 using the instructions I mentioned above and this weekend tried out Frigate 0.17 Beta 1 and it's working well.

@atrozgrifo commented on GitHub (Jan 12, 2026): I was running Frigate 0.16 using the instructions I mentioned above and this weekend tried out Frigate 0.17 Beta 1 and it's working well.

kerem commented

Author

Owner

@LeidenSpain commented on GitHub (Jan 12, 2026):

The method that has worked perfectly for me is creating a privileged LXC with the Docker script, and then running Frigate's compose.

It's working perfectly.

@LeidenSpain commented on GitHub (Jan 12, 2026): The method that has worked perfectly for me is creating a privileged LXC with the Docker script, and then running Frigate's compose. It's working perfectly.

kerem commented

Author

Owner

@samwathegreat commented on GitHub (Jan 12, 2026):

Unprivileged docker LXC + official docker image worked great for me. But
I'm not using any kind of detection (disabled on every camera) and only
doing 24/7 recording. So this is an option if you don't need detection (or
even if you plan to use CPU detection).

On Mon, Jan 12, 2026 at 3:07 PM LeidenSpain @.***>
wrote:

LeidenSpain left a comment (community-scripts/ProxmoxVE#6915)
https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3740293252

The method that has worked perfectly for me is creating a privileged LXC
with the Docker script, and then running Frigate's compose.

It's working perfectly.

—
Reply to this email directly, view it on GitHub
https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3740293252,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/ABEXEGHY4RE5LDKXIWMUJTL4GP5H7AVCNFSM6AAAAACEC6ZHIWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTONBQGI4TGMRVGI
.
You are receiving this because you commented.Message ID:
@.***>

@samwathegreat commented on GitHub (Jan 12, 2026): Unprivileged docker LXC + official docker image worked great for me. But I'm not using any kind of detection (disabled on every camera) and only doing 24/7 recording. So this is an option if you don't need detection (or even if you plan to use CPU detection). On Mon, Jan 12, 2026 at 3:07 PM LeidenSpain ***@***.***> wrote: > *LeidenSpain* left a comment (community-scripts/ProxmoxVE#6915) > <https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3740293252> > > The method that has worked perfectly for me is creating a privileged LXC > with the Docker script, and then running Frigate's compose. > > It's working perfectly. > > — > Reply to this email directly, view it on GitHub > <https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3740293252>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ABEXEGHY4RE5LDKXIWMUJTL4GP5H7AVCNFSM6AAAAACEC6ZHIWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTONBQGI4TGMRVGI> > . > You are receiving this because you commented.Message ID: > ***@***.***> >

kerem commented

Author

Owner

@Adrien40 commented on GitHub (Jan 13, 2026):

Hello :
mini PC Intel N150, Google Coral USB, Proxmox 9.1.4

It worked a few days ago

✔️ Installed Dependencies
✖️ in line 1841: exit code 100 (APT: Package manager error (broken packages / dependency problems))
→ apt install -y /tmp/tmp.SiDbvLzf28/intel-igc-core-2_2.27.10%2B20617_amd64.deb
--- Last 10 lines of silent log ---
requested an impossible situation or if you are using the unstable
distribution that some required packages have not yet been created
or been moved out of Incoming.
The following information may help to resolve the situation:

The following packages have unmet dependencies:
intel-igc-core-2 : Depends: libc6 (>= 2.38) but 2.36-9+deb12u13 is to be installed
Depends: libstdc++6 (>= 13) but 12.2.0-14+deb12u1 is to be installed
Depends: libzstd1 (>= 1.5.5) but 1.5.4+dfsg2-5 is to be installed
E: Unable to correct problems, you have held broken packages.

📋 View full log (2702 lines): /root/.install-024304e9.log
✖️ Installation failed in container 102 (exit code: 100)

✔ Installation log: /tmp/install-lxc-102-024304e9.log

Remove broken container 102? (Y/n) [auto-remove in 60s]:
No response - auto-removing container
Removing container 102
✔️ Container 102 removed

@Adrien40 commented on GitHub (Jan 13, 2026): Hello : mini PC Intel N150, Google Coral USB, Proxmox 9.1.4 It worked a few days ago ✔️ Installed Dependencies ✖️ in line 1841: exit code 100 (APT: Package manager error (broken packages / dependency problems)) → apt install -y /tmp/tmp.SiDbvLzf28/intel-igc-core-2_2.27.10%2B20617_amd64.deb --- Last 10 lines of silent log --- requested an impossible situation or if you are using the unstable distribution that some required packages have not yet been created or been moved out of Incoming. The following information may help to resolve the situation: The following packages have unmet dependencies: intel-igc-core-2 : Depends: libc6 (>= 2.38) but 2.36-9+deb12u13 is to be installed Depends: libstdc++6 (>= 13) but 12.2.0-14+deb12u1 is to be installed Depends: libzstd1 (>= 1.5.5) but 1.5.4+dfsg2-5 is to be installed E: Unable to correct problems, you have held broken packages. ----------------------------------- 📋 View full log (2702 lines): /root/.install-024304e9.log ✖️ Installation failed in container 102 (exit code: 100) ✔ Installation log: /tmp/install-lxc-102-024304e9.log Remove broken container 102? (Y/n) [auto-remove in 60s]: No response - auto-removing container Removing container 102 ✔️ Container 102 removed

kerem commented

Author

Owner

@Patrick010 commented on GitHub (Jan 13, 2026):

Wrong debian version? Try latest or testing.

@Patrick010 commented on GitHub (Jan 13, 2026): Wrong debian version? Try latest or testing.

kerem commented

Author

Owner

@SuperSpe commented on GitHub (Jan 30, 2026):

I am searching for a 0.16 script too.
The fact that you guys install frigate inside a docker container, inside a lxc, inside proxmox, inside a host is very inconvenient in terms of Virtualization layer stratification, but if it works, it works.

@SuperSpe commented on GitHub (Jan 30, 2026): I am searching for a 0.16 script too. The fact that you guys install frigate inside a docker container, inside a lxc, inside proxmox, inside a host is very inconvenient in terms of Virtualization layer stratification, but if it works, it works.

kerem commented

Author

Owner

@SuperSpe commented on GitHub (Jan 30, 2026):

Hi,

I created a script based on the previous LXC Frigate Scripts. I was able to update all the reference to the lates mentioned in the dockerfile and have successfully tested this for installing (frigate 16.0) and updating to 16.3 in a fresh lxc container. I am succesfully using this with a coral USB accelerator. Maybe this helps others.

install_frigate.sh

I'll give you try. Is it a wizard for the creation of the lxc or do I need to run it in a blank debian VM?

@SuperSpe commented on GitHub (Jan 30, 2026): > Hi, > > I created a script based on the previous LXC Frigate Scripts. I was able to update all the reference to the lates mentioned in the dockerfile and have successfully tested this for installing (frigate 16.0) and updating to 16.3 in a fresh lxc container. I am succesfully using this with a coral USB accelerator. Maybe this helps others. > > [install_frigate.sh](https://github.com/user-attachments/files/24540598/install_frigate.sh) I'll give you try. Is it a wizard for the creation of the lxc or do I need to run it in a blank debian VM?

kerem commented

Author

Owner

@croessi commented on GitHub (Feb 1, 2026):

There is an first preview, but it doesnt work fine at the moment. The annoying thing about Frigate is that it takes ages to build, and the slightest mistake means you have to start all over again. Unfortunately, that takes up a huge amount of time.

bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVED/main/ct/frigate.sh)"

@MickLesk Do you still face issues? If yes maybe my script is of help. It is running fine on Debian12, priviledged container with GPU and CORAL support. Even updating seems to work fine.

install_frigate.sh

@croessi commented on GitHub (Feb 1, 2026): > There is an first preview, but it doesnt work fine at the moment. The annoying thing about Frigate is that it takes ages to build, and the slightest mistake means you have to start all over again. Unfortunately, that takes up a huge amount of time. > > bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVED/main/ct/frigate.sh)" @MickLesk Do you still face issues? If yes maybe my script is of help. It is running fine on Debian12, priviledged container with GPU and CORAL support. Even updating seems to work fine. [install_frigate.sh](https://github.com/user-attachments/files/24991983/install_frigate.sh)

kerem commented

Author

Owner

@wlad1164 commented on GitHub (Feb 1, 2026):

@croessi Hi! Am I correct in understanding that I need to create a clean container and then run your script inside it?

@wlad1164 commented on GitHub (Feb 1, 2026): @croessi Hi! Am I correct in understanding that I need to create a clean container and then run your script inside it?

kerem commented

Author

Owner

@croessi commented on GitHub (Feb 1, 2026):

@wlad1164 Yes just create a Debian 12 (bookworm) container. I tested with privileged, 4 cores, 4GB RAM, 20GB Disk and GPU passthrough. You may use this script on the proxmox root to create it:

debian12.sh

@croessi commented on GitHub (Feb 1, 2026): @wlad1164 Yes just create a Debian 12 (bookworm) container. I tested with privileged, 4 cores, 4GB RAM, 20GB Disk and GPU passthrough. You may use this script on the proxmox root to create it: [debian12.sh](https://github.com/user-attachments/files/24992163/debian12.sh)

kerem commented

Author

Owner

@wlad1164 commented on GitHub (Feb 1, 2026):

@croessi In my case, the script crashes here:

cp dist/vec0.so /usr/local/lib
Stopping go2rtc.service (if running)...
Killing any remaining go2rtc processes...
Terminated
root@frigate16:~#

This is line 244 of the script.

@wlad1164 commented on GitHub (Feb 1, 2026): @croessi In my case, the script crashes here: + cp dist/vec0.so /usr/local/lib Stopping go2rtc.service (if running)... Killing any remaining go2rtc processes... Terminated root@frigate16:~# This is line 244 of the script.

kerem commented

Author

Owner

@wlad1164 commented on GitHub (Feb 1, 2026):

@croessi I replaced part of the install_go2rtc function

echo "Terminating go2rtc processes..."
GO2RTC_PID=$(pgrep -f "/usr/local/go2rtc/bin/go2rtc" 2>/dev/null)

if [ -n "$GO2RTC_PID" ]; then
    kill "$GO2RTC_PID" 2>/dev/null || true
    sleep 1
    if kill -0 "$GO2RTC_PID" 2>/dev/null; then
        kill -9 "$GO2RTC_PID" 2>/dev/null || true
    fi
fi

  tmp="/usr/local/go2rtc/bin/go2rtc.new"
  echo "Downloading go2rtc..."
  wget -q --progress=bar:force -O "${tmp}" "${url}"
  
  if [ $? -eq 0 ]; then
      chmod 755 "${tmp}"
      mv "${tmp}" /usr/local/go2rtc/bin/go2rtc
      echo "Installation completed"
  else
      echo "Download failed"
      rm -f "${tmp}"
      exit 1
  fi

and was able to move forward. Let's see if I can get to the end.

@wlad1164 commented on GitHub (Feb 1, 2026): @croessi I replaced part of the install_go2rtc function ``` echo "Terminating go2rtc processes..." GO2RTC_PID=$(pgrep -f "/usr/local/go2rtc/bin/go2rtc" 2>/dev/null) if [ -n "$GO2RTC_PID" ]; then kill "$GO2RTC_PID" 2>/dev/null || true sleep 1 if kill -0 "$GO2RTC_PID" 2>/dev/null; then kill -9 "$GO2RTC_PID" 2>/dev/null || true fi fi tmp="/usr/local/go2rtc/bin/go2rtc.new" echo "Downloading go2rtc..." wget -q --progress=bar:force -O "${tmp}" "${url}" if [ $? -eq 0 ]; then chmod 755 "${tmp}" mv "${tmp}" /usr/local/go2rtc/bin/go2rtc echo "Installation completed" else echo "Download failed" rm -f "${tmp}" exit 1 fi ``` and was able to move forward. Let's see if I can get to the end.

kerem commented

Author

Owner

@wlad1164 commented on GitHub (Feb 1, 2026):

I'm stuck here at the moment... I can't figure out how to fix it...

+ apt-get install --no-install-recommends --no-install-suggests -y i965-va-driver-shaders
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
Package i965-va-driver-shaders is not available, but is referred to by another package.
This may mean that the package is missing, has been obsoleted, or
is only available from another source

E: Package 'i965-va-driver-shaders' has no installation candidate
+ apt-get -qq install --no-install-recommends --no-install-suggests -y intel-gpu-tools onevpl-tools libva-drm2 mesa-va-drivers radeontop
Selecting previously unselected package libxv1:amd64.
(Reading database ... 58290 files and directories currently installed.)
Preparing to unpack .../libxv1_2%3a1.0.11-1.1_amd64.deb ...
Unpacking libxv1:amd64 (2:1.0.11-1.1) ...
Selecting previously unselected package intel-gpu-tools.
Preparing to unpack .../intel-gpu-tools_1.27.1-1_amd64.deb ...
Unpacking intel-gpu-tools (1.27.1-1) ...
Selecting previously unselected package libvpl2.
Preparing to unpack .../libvpl2_2023.1.1-1_amd64.deb ...
Unpacking libvpl2 (2023.1.1-1) ...
Selecting previously unselected package onevpl-tools.
Preparing to unpack .../onevpl-tools_2023.1.1-1_amd64.deb ...
Unpacking onevpl-tools (2023.1.1-1) ...
Selecting previously unselected package radeontop.
Preparing to unpack .../radeontop_1.4-2_amd64.deb ...
Unpacking radeontop (1.4-2) ...
Setting up libvpl2 (2023.1.1-1) ...
Setting up onevpl-tools (2023.1.1-1) ...
Setting up libxv1:amd64 (2:1.0.11-1.1) ...
Setting up radeontop (1.4-2) ...
Setting up intel-gpu-tools (1.27.1-1) ...
Processing triggers for man-db (2.11.2-2) ...
Processing triggers for libc-bin (2.36-9+deb12u13) ...
+ apt-get install -y dpkg
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
dpkg is already the newest version (1.21.22).
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
+ wget -qO - https://repositories.intel.com/gpu/intel-graphics.key
+ gpg --yes --dearmor --output /usr/share/keyrings/intel-graphics.gpg
gpg: no valid OpenPGP data found.

@wlad1164 commented on GitHub (Feb 1, 2026): I'm stuck here at the moment... I can't figure out how to fix it... ``` + apt-get install --no-install-recommends --no-install-suggests -y i965-va-driver-shaders Reading package lists... Done Building dependency tree... Done Reading state information... Done Package i965-va-driver-shaders is not available, but is referred to by another package. This may mean that the package is missing, has been obsoleted, or is only available from another source E: Package 'i965-va-driver-shaders' has no installation candidate + apt-get -qq install --no-install-recommends --no-install-suggests -y intel-gpu-tools onevpl-tools libva-drm2 mesa-va-drivers radeontop Selecting previously unselected package libxv1:amd64. (Reading database ... 58290 files and directories currently installed.) Preparing to unpack .../libxv1_2%3a1.0.11-1.1_amd64.deb ... Unpacking libxv1:amd64 (2:1.0.11-1.1) ... Selecting previously unselected package intel-gpu-tools. Preparing to unpack .../intel-gpu-tools_1.27.1-1_amd64.deb ... Unpacking intel-gpu-tools (1.27.1-1) ... Selecting previously unselected package libvpl2. Preparing to unpack .../libvpl2_2023.1.1-1_amd64.deb ... Unpacking libvpl2 (2023.1.1-1) ... Selecting previously unselected package onevpl-tools. Preparing to unpack .../onevpl-tools_2023.1.1-1_amd64.deb ... Unpacking onevpl-tools (2023.1.1-1) ... Selecting previously unselected package radeontop. Preparing to unpack .../radeontop_1.4-2_amd64.deb ... Unpacking radeontop (1.4-2) ... Setting up libvpl2 (2023.1.1-1) ... Setting up onevpl-tools (2023.1.1-1) ... Setting up libxv1:amd64 (2:1.0.11-1.1) ... Setting up radeontop (1.4-2) ... Setting up intel-gpu-tools (1.27.1-1) ... Processing triggers for man-db (2.11.2-2) ... Processing triggers for libc-bin (2.36-9+deb12u13) ... + apt-get install -y dpkg Reading package lists... Done Building dependency tree... Done Reading state information... Done dpkg is already the newest version (1.21.22). 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. + wget -qO - https://repositories.intel.com/gpu/intel-graphics.key + gpg --yes --dearmor --output /usr/share/keyrings/intel-graphics.gpg gpg: no valid OpenPGP data found. ```

kerem commented

Author

Owner

@vovushko commented on GitHub (Feb 1, 2026):

@croessi Hi! I had errors installing the script:
tar: ssdlite_mobilenet_v2_coco_2018_05_09/checkpoint: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09/model.ckpt.data-00000-of-00001: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09/model.ckpt.meta: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09/model.ckpt.index: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09/saved_model/saved_model.pb: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09/pipeline.config: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09/frozen_inference_graph.pb: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09/saved_model/variables: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09/saved_model: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: Exiting with failure status due to previous errors

@vovushko commented on GitHub (Feb 1, 2026): @croessi Hi! I had errors installing the script: `tar: ssdlite_mobilenet_v2_coco_2018_05_09/checkpoint: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09/model.ckpt.data-00000-of-00001: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09/model.ckpt.meta: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09/model.ckpt.index: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09/saved_model/saved_model.pb: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09/pipeline.config: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09/frozen_inference_graph.pb: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09/saved_model/variables: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09/saved_model: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: ssdlite_mobilenet_v2_coco_2018_05_09: Cannot change ownership to uid 345018, gid 89939: Invalid argument tar: Exiting with failure status due to previous errors `

kerem commented

Author

Owner

@croessi commented on GitHub (Feb 1, 2026):

@wlad1164 & @vovushko Have you tried to run it with a fresh container generated with this script debian12.sh? You need to install it in advanced mode with the parameters:

Privileged
4GB RAM
4 cores
20GB disk

@croessi commented on GitHub (Feb 1, 2026): @wlad1164 & @vovushko Have you tried to run it with a fresh container generated with this script [debian12.sh](https://github.com/user-attachments/files/24992954/debian12.sh)? You need to install it in advanced mode with the parameters: - Privileged - 4GB RAM - 4 cores - 20GB disk

kerem commented

Author

Owner

@MickLesk commented on GitHub (Feb 1, 2026):

Please provide this into seperate issue or in the fork. Its Not related to us.

@MickLesk commented on GitHub (Feb 1, 2026): Please provide this into seperate issue or in the fork. Its Not related to us.

kerem commented

Author

Owner

@vovushko commented on GitHub (Feb 1, 2026):

@wlad1164 & @vovushko Have you tried to run it with a fresh container generated with this script debian12.sh? You need to install it in advanced mode with the parameters:

Privileged

4GB RAM

4 cores

20GB disk

yes

@vovushko commented on GitHub (Feb 1, 2026): > [@wlad1164](https://github.com/wlad1164) & [@vovushko](https://github.com/vovushko) Have you tried to run it with a fresh container generated with this script [debian12.sh](https://github.com/user-attachments/files/24992954/debian12.sh)? You need to install it in advanced mode with the parameters: > > * Privileged > * 4GB RAM > * 4 cores > * 20GB disk yes

kerem commented

Author

Owner

@wlad1164 commented on GitHub (Feb 1, 2026):

@wlad1164 & @vovushko Have you tried to run it with a fresh container generated with this script debian12.sh? You need to install it in advanced mode with the parameters:

Privileged

4GB RAM

4 cores

20GB disk

Yes

@wlad1164 commented on GitHub (Feb 1, 2026): > [@wlad1164](https://github.com/wlad1164) & [@vovushko](https://github.com/vovushko) Have you tried to run it with a fresh container generated with this script [debian12.sh](https://github.com/user-attachments/files/24992954/debian12.sh)? You need to install it in advanced mode with the parameters: > > * Privileged > * 4GB RAM > * 4 cores > * 20GB disk Yes

kerem commented

Author

Owner

@wlad1164 commented on GitHub (Feb 1, 2026):

@croessi @vovushko I was able to complete the installation! I had to modify the script in several places, but now it works. I installed version 0.16.4-4131252a. I had to add "non-free non-free-firmware" to sudo nano /etc/apt/sources.list to install the "i965-va-driver-shaders" package, as well as completely disable the "repositories.intel.com" repositories and remove the "intel-level-zero-gpu" and "libmfxgen1" packages.
These are the main changes I had to make. There were a few other minor ones, but I don't remember all the details.

@wlad1164 commented on GitHub (Feb 1, 2026): @croessi @vovushko I was able to complete the installation! I had to modify the script in several places, but now it works. I installed version 0.16.4-4131252a. I had to add "non-free non-free-firmware" to sudo nano /etc/apt/sources.list to install the "i965-va-driver-shaders" package, as well as completely disable the "repositories.intel.com" repositories and remove the "intel-level-zero-gpu" and "libmfxgen1" packages. These are the main changes I had to make. There were a few other minor ones, but I don't remember all the details.

kerem commented

https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3831126194

Author

Owner

@MickLesk commented on GitHub (Feb 1, 2026):

I lock it now.

@MickLesk commented on GitHub (Feb 1, 2026): I lock it now. https://github.com/community-scripts/ProxmoxVE/issues/6915#issuecomment-3831126194

kerem commented