[GH-ISSUE #2239] CS1.6 LGSM Script suddenly stopped working in some, but not all, cases. #1706

New issue

Closed

opened 2026-02-27 02:58:38 +03:00 by kerem · 3 comments

kerem commented

2026-02-27 02:58:38 +03:00

Owner

Originally created by @rip1980 on GitHub (Mar 23, 2019).
Original GitHub issue: https://github.com/GameServerManagers/LinuxGSM/issues/2239

(Posted here because I think this is a LGSM issue. Something, maybe an APT update, broke it. I doubt I'm special no matter what my mom says. Others may, or may soon, experience an issue if some dependency has broken the script but I can not locate it.)

User Story

5:30-sh this morning started having an issue where some ports/game instances are not working. Box is running several game (CS 1.6) servers on ports 27015 to 27022, and in shell, these are running and things are happening (logs show bots killing each other, map changes, worlds spawning, etc.) Only games on 27016, 27020 and 27022 are connecting and reachable on the outside internet, although all show running internally.

Netstat show them running, the I noticed the script in python is flapping for the broken servers (detects server NOT running and restarts them, HOWEVER, I think they did not crash, I spent hours going through logs and really don't have a smoking gun error.) I have not figured out the delta between working and not working instances.

TS things: LOGS LOGS LOGS, nothing stands out. Even apt updates happened after it broke, so probably not the cause. FW rule to allow anything just in case. Running as debug stops the flapping (script restarting the game server but it's still not reachable outside (appears for a split second then dies externally) Lots of soft and box restarts. updated games, updated lgsm, forced validation checks, etc.

I've run servers for years, migrated to this physical machine in Sept, but ran it for a few years at the same host before changing boxes. Linux user since .99beta16, I'm usually OK at this stuff. :( This has me stumped. Not sure if it's a script thing or a Steam/CS issue, literally nothing had changed. Only time I've seen something like this is when liblist.gam gets overwritten, and it's OK in this case.

TODO: Possible roll old backup over broken instances and see if it works. Spend more time looking at date stamps to try to see what changed.

https://hastebin.com/dowolekami

Basic info

Distro: Ubuntu 18.04.2 LTS
Arch: x86_64
Kernel: 4.15.0-46-generic
Hostname: dedibox
tmux: tmux 2.6
glibc: 2.27
CS1.6, multiple servers on ports 27015 to 27022

Further Information

Been running LGSM for a few years, this setup has been running about 7 months with no real changes (Occasional APT updates, Steam updates,) No known changes to correlate with this issue (not an update then it broke type issue.)

My notes and captures: https://hastebin.com/dowolekami

To Reproduce

Work on 3 instances, does not work on 5 others. I have not identified the delta between works vs doesn't. Suspect script itself is having an issue not reading the port state correctly.

Maybe an earlier python or os update tripped it up? However, current python version is not bleeding edge:
csserver1@dedibox:~/serverfiles/cstrike/logs$ python
Python 2.7.15rc1 (default, Nov 12 2018, 14:31:15)
[GCC 7.3.0] on linux2

Originally created by @rip1980 on GitHub (Mar 23, 2019). Original GitHub issue: https://github.com/GameServerManagers/LinuxGSM/issues/2239 (Posted here because I think this is a LGSM issue. Something, maybe an APT update, broke it. I doubt I'm special no matter what my mom says. Others may, or may soon, experience an issue if some dependency has broken the script but I can not locate it.) ## User Story 5:30-sh this morning started having an issue where some ports/game instances are not working. Box is running several game (CS 1.6) servers on ports 27015 to 27022, and in shell, these are running and things are happening (logs show bots killing each other, map changes, worlds spawning, etc.) Only games on 27016, 27020 and 27022 are connecting and reachable on the outside internet, although all show running internally. Netstat show them running, the I noticed the script in python is flapping for the broken servers (detects server NOT running and restarts them, HOWEVER, I think they did not crash, I spent hours going through logs and really don't have a smoking gun error.) I have not figured out the delta between working and not working instances. TS things: LOGS LOGS LOGS, nothing stands out. Even apt updates happened after it broke, so probably not the cause. FW rule to allow anything just in case. Running as debug stops the flapping (script restarting the game server but it's still not reachable outside (appears for a split second then dies externally) Lots of soft and box restarts. updated games, updated lgsm, forced validation checks, etc. I've run servers for years, migrated to this physical machine in Sept, but ran it for a few years at the same host before changing boxes. Linux user since .99beta16, I'm usually OK at this stuff. :( This has me stumped. Not sure if it's a script thing or a Steam/CS issue, literally nothing had changed. Only time I've seen something like this is when liblist.gam gets overwritten, and it's OK in this case. TODO: Possible roll old backup over broken instances and see if it works. Spend more time looking at date stamps to try to see what changed. https://hastebin.com/dowolekami ## Basic info Distro: Ubuntu 18.04.2 LTS Arch: x86_64 Kernel: 4.15.0-46-generic Hostname: dedibox tmux: tmux 2.6 glibc: 2.27 CS1.6, multiple servers on ports 27015 to 27022 ## Further Information Been running LGSM for a few years, this setup has been running about 7 months with no real changes (Occasional APT updates, Steam updates,) No known changes to correlate with this issue (not an update then it broke type issue.) My notes and captures: https://hastebin.com/dowolekami ## To Reproduce Work on 3 instances, does not work on 5 others. I have not identified the delta between works vs doesn't. Suspect script itself is having an issue not reading the port state correctly. Maybe an earlier python or os update tripped it up? However, current python version is not bleeding edge: csserver1@dedibox:~/serverfiles/cstrike/logs$ python Python 2.7.15rc1 (default, Nov 12 2018, 14:31:15) [GCC 7.3.0] on linux2

kerem closed this issue

2026-02-27 02:58:39 +03:00

kerem commented

2026-02-27 02:58:39 +03:00

Author

Owner

@rip1980 commented on GitHub (Mar 23, 2019):

Just an update: A backup copy of the whole server worked IN ONE CASE. Literally just blew the old backup on and it launched and worked. It's late here, tomorrow I will try to do a compare between the broken and working scripts. Backup is a few months old, so hoping the deltas are not too much. Thanks in advance for any ideas.

In another case, rolled back another server from the same day (Remember running 8 servers on different ports on a dedicated box, all CS1.6 with different mods [Gungame, Deathmatch, etc., nothing too special.]) and it still fails. Tried it before allowing the game or script to update, and after. Originally blew the backup on and it failed.

So literally, a working copy from months ago worked in one case, and did not in another. Same box, backups were made the same day, works in one case and not in another. Simply gsquery shows as failed even though dt shows it running.

No changes prior to issues. This is hurting my brain because it make no sense. 8}

@rip1980 commented on GitHub (Mar 23, 2019): Just an update: A backup copy of the whole server worked IN ONE CASE. Literally just blew the old backup on and it launched and worked. It's late here, tomorrow I will try to do a compare between the broken and working scripts. Backup is a few months old, so hoping the deltas are not too much. Thanks in advance for any ideas. In another case, rolled back another server from the same day (Remember running 8 servers on different ports on a dedicated box, all CS1.6 with different mods [Gungame, Deathmatch, etc., nothing too special.]) and it still fails. Tried it before allowing the game or script to update, and after. Originally blew the backup on and it failed. So literally, a working copy from months ago worked in one case, and did not in another. Same box, backups were made the same day, works in one case and not in another. Simply gsquery shows as failed even though dt shows it running. No changes prior to issues. This is hurting my brain because it make no sense. 8}

kerem commented

2026-02-27 02:58:39 +03:00

Author

Owner

@rip1980 commented on GitHub (Mar 25, 2019):

Closing issue. Install did not work, deleting EVERYTHING and wget a new script did work. Something broke badly but 100% LGSM does work, so, shrug. Latest script works so can't say it's LGSM itself.

@rip1980 commented on GitHub (Mar 25, 2019): Closing issue. Install did not work, deleting EVERYTHING and wget a new script did work. Something broke badly but 100% LGSM does work, so, shrug. Latest script works so can't say it's LGSM itself.

kerem commented

2026-02-27 02:58:39 +03:00

Author

Owner

@lock[bot] commented on GitHub (Mar 27, 2020):

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.

@lock[bot] commented on GitHub (Mar 27, 2020): This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.