starred/LinuxGSM

Fork 0

mirror of https://github.com/GameServerManagers/LinuxGSM.git synced 2026-04-25 22:25:59 +03:00

[GH-ISSUE #807] [Rust] Admin Restart Command In Game Does Not Restart #648

New issue

Open

opened 2026-02-27 02:02:52 +03:00 by kerem · 7 comments

kerem commented

2026-02-27 02:02:52 +03:00

Owner

Originally created by @bedoherty on GitHub (May 6, 2016).
Original GitHub issue: https://github.com/GameServerManagers/LinuxGSM/issues/807

Originally assigned to: @UltimateByte on GitHub.

When the admin in game does a restart command, the LGSM Rust server restarts, which first stops the server, but then it never starts back up. A fix for this would be incredibly helpful.

Originally created by @bedoherty on GitHub (May 6, 2016). Original GitHub issue: https://github.com/GameServerManagers/LinuxGSM/issues/807 Originally assigned to: @UltimateByte on GitHub. When the admin in game does a restart command, the LGSM Rust server restarts, which first stops the server, but then it never starts back up. A fix for this would be incredibly helpful.

kerem added the

type: feature

game: Rust

labels

2026-02-27 02:02:52 +03:00

kerem commented

2026-02-27 02:02:52 +03:00

Author

Owner

@UltimateByte commented on GitHub (May 6, 2016):

Isolated issue. More details required.

Edit : I get what you're saying. Don't use that command. Use LGSM restart function. Optionally, use the rcon "save" command before restarting. Lastly, why would one manually restart a rust server?

@UltimateByte commented on GitHub (May 6, 2016): Isolated issue. More details required. Edit : I get what you're saying. Don't use that command. Use LGSM restart function. Optionally, use the rcon "save" command before restarting. Lastly, why would one manually restart a rust server?

kerem commented

2026-02-27 02:02:52 +03:00

Author

Owner

@aelor commented on GitHub (Jul 29, 2017):

Restarting manually or automatically is upto the server owner. What matters is the server restart messaging. When we do the global.restart , the messaging is quite proper with countdown which doesn't happen in case of ./rustserver restart. The server just abruptly stops and restarts after sometime. Instead of this abrupt stop if a messaging could be sent to the ingame players similar to global.restart this command would be much more handy

@aelor commented on GitHub (Jul 29, 2017): Restarting manually or automatically is upto the server owner. What matters is the server restart messaging. When we do the `global.restart `, the messaging is quite proper with countdown which doesn't happen in case of `./rustserver restart`. The server just abruptly stops and restarts after sometime. Instead of this abrupt stop if a messaging could be sent to the ingame players similar to `global.restart` this command would be much more handy

kerem commented

2026-02-27 02:02:52 +03:00

Author

Owner

@UltimateByte commented on GitHub (Jul 30, 2017):

@aelor If you find a way to send RCON commands into Rust, then please share. Otherwise what you proposed which is what we all want remains undoable.

@UltimateByte commented on GitHub (Jul 30, 2017): @aelor If you find a way to send RCON commands into Rust, then please share. Otherwise what you proposed which is what we all want remains undoable.

kerem commented

2026-02-27 02:02:52 +03:00

Author

Owner

@lotkao commented on GitHub (Dec 3, 2017):

@UltimateByte The great thing about doing "global.restart" via RCON is that it does an automatic save right before shutting down the server. Also, when people get disconnected they see Disconnected: Server Restarting at the bottom of the "Server Details" that they get thrown back to, and in the Console it says:

Disconnect Reason: Server Restarting
Disconnected (Server Restarting) - returning to main menu

When doing rustserver restart people who get disconnected see Disconnected: Timed Out at the bottom of the "Server Details" that they get thrown back to, and in the Console it says:

Disconnected (Timed Out) - returning to main menu

The only problem with using global.restart is that a server that was started with LinuxGSM doesn't actually get restarted and needs to be manually started again.

One way to send RCON commands into Rust may be https://github.com/websockets/wscat

I was able to connect to RCON with: wscat -c ws://SERVER_IP:PORT/PASSWORD

And I successfully send the restart command via:
{"Identifier":1,Message:"global.restart",Name:"WebRcon"}

@lotkao commented on GitHub (Dec 3, 2017): @UltimateByte The great thing about doing "global.restart" via RCON is that it does an automatic save right before shutting down the server. Also, when people get disconnected they see `Disconnected: Server Restarting` at the bottom of the "Server Details" that they get thrown back to, and in the Console it says: ``` Disconnect Reason: Server Restarting Disconnected (Server Restarting) - returning to main menu ``` When doing `rustserver restart` people who get disconnected see `Disconnected: Timed Out` at the bottom of the "Server Details" that they get thrown back to, and in the Console it says: ``` Disconnected (Timed Out) - returning to main menu ``` The only problem with using `global.restart` is that a server that was started with LinuxGSM doesn't actually get restarted and needs to be manually started again. One way to send RCON commands into Rust may be https://github.com/websockets/wscat I was able to connect to RCON with: `wscat -c ws://SERVER_IP:PORT/PASSWORD` And I successfully send the restart command via: `{"Identifier":1,Message:"global.restart",Name:"WebRcon"}`

kerem commented

2026-02-27 02:02:52 +03:00

Author

Owner

@UltimateByte commented on GitHub (Dec 3, 2017):

Now we're talking.

@UltimateByte commented on GitHub (Dec 3, 2017): Now we're talking.

kerem commented

2026-02-27 02:02:53 +03:00

Author

Owner

@lotkao commented on GitHub (Dec 4, 2017):

I've done some modifications to the wscat, so it can be given the json-string when being called.

https://github.com/gitforwebwork/wscat/tree/addexecute

Usage:

wscat -c ws://SERVER_IP:PORT/PASSWORD -x '{"Identifier":1,Message:"global.restart",Name:"WebRcon"}'

I added one more parameter do deal with slow requests, but the default works fine for me.

-w or --wait <seconds> will make it wait for a set amount of seconds before quitting (default is 2 seconds)

@lotkao commented on GitHub (Dec 4, 2017): I've done some modifications to the wscat, so it can be given the json-string when being called. https://github.com/gitforwebwork/wscat/tree/addexecute Usage: `wscat -c ws://SERVER_IP:PORT/PASSWORD -x '{"Identifier":1,Message:"global.restart",Name:"WebRcon"}'` I added one more parameter do deal with slow requests, but the default works fine for me. `-w or --wait <seconds>` will make it wait for a set amount of seconds before quitting (default is 2 seconds)

kerem commented

2026-02-27 02:02:53 +03:00

Author

Owner

@lotkao commented on GitHub (Dec 4, 2017):

For now, I have created a little script that I use to monitor the server. I know it's crude, but it does the job. Using the script in a cronjob, I can do `global.restart' via RCON and the server comes back up automatically.

I saved the script to /home/rustserver/monitor-rustserver.sh

#! /bin/bash
rustserver="/home/rustserver/"
rusthash=".cache/servermonitoringhash"

# delete too old hash (a badly timed server crash/reboot could cause that)
if [ -f ${rustserver}${rusthash} ]
then
    hashage=$((`date +%s` - `date -r ${rustserver}${rusthash} +%s`))
    if [ $hashage -gt 120 ]
    then
        rm ${rustserver}${rusthash}
    fi
fi

#start rustserver, if no process exists
if [ $(pidof RustDedicated | wc -l) -eq 0 ] && [ ! -f ${rustserver}${rusthash} ]
then
    touch ${rustserver}${rusthash}
    echo "$(date)"
    ${rustserver}rustserver start
    rm ${rustserver}${rusthash}

# monitor rustserver, if it is running for at least 3 minutes
elif [ $(pidof RustDedicated | wc -l) -eq 1 ] && [ $(ps -o etimes= -p $(pidof RustDedicated) | xargs) -gt 180 ] && [ ! -f ${rustserver}${rusthash} ]
then
    touch ${rustserver}${rusthash}
    echo "$(date)"
    ${rustserver}rustserver monitor
    rm ${rustserver}${rusthash}
fi

I made it wait for 180 seconds, before doing rustserver monitor, because in some "unfortunately timed" tests the server did not come back up fast enough and it restarted it again.

The cronjobs I am running to monitor and restart the server:

* * * * * /home/rustserver/monitor-rustserver.sh >> /home/rustserver/log/uptime/monitor-rustserver.sh 2>&1
* * * * * sleep 10; /home/rustserver/monitor-rustserver.sh >> /home/rustserver/log/uptime/monitor-rustserver.log 2>&1
* * * * * sleep 20; /home/rustserver/monitor-rustserver.sh >> /home/rustserver/log/uptime/monitor-rustserver.log 2>&1
* * * * * sleep 30; /home/rustserver/monitor-rustserver.sh >> /home/rustserver/log/uptime/monitor-rustserver.log 2>&1
* * * * * sleep 40; /home/rustserver/monitor-rustserver.sh >> /home/rustserver/log/uptime/monitor-rustserver.log 2>&1
* * * * * sleep 50; /home/rustserver/monitor-rustserver.sh >> /home/rustserver/log/uptime/monitor-rustserver.log 2>&1

0 5 * * * wscat -c ws://SERVER_IP:PORT/PASSWORD -x '{"Identifier":1,Message:"global.restart 10",Name:"WebRcon"}' > /dev/null 2>&1

If you want to use this... make sure you have the folders /home/rustserver/.cache/ and /home/rustserver/log/uptime/, or modify the script and cronjobs to use whatever folders you want.

Even though the rusthash should prevent the script to run rustserver start or rustserver monitor twice, I would not recommend running the monitoring script every second.

@lotkao commented on GitHub (Dec 4, 2017): For now, I have created a little script that I use to monitor the server. I know it's crude, but it does the job. Using the script in a cronjob, I can do `global.restart' via RCON and the server comes back up automatically. I saved the script to /home/rustserver/monitor-rustserver.sh ``` #! /bin/bash rustserver="/home/rustserver/" rusthash=".cache/servermonitoringhash" # delete too old hash (a badly timed server crash/reboot could cause that) if [ -f ${rustserver}${rusthash} ] then hashage=$((`date +%s` - `date -r ${rustserver}${rusthash} +%s`)) if [ $hashage -gt 120 ] then rm ${rustserver}${rusthash} fi fi #start rustserver, if no process exists if [ $(pidof RustDedicated | wc -l) -eq 0 ] && [ ! -f ${rustserver}${rusthash} ] then touch ${rustserver}${rusthash} echo "$(date)" ${rustserver}rustserver start rm ${rustserver}${rusthash} # monitor rustserver, if it is running for at least 3 minutes elif [ $(pidof RustDedicated | wc -l) -eq 1 ] && [ $(ps -o etimes= -p $(pidof RustDedicated) | xargs) -gt 180 ] && [ ! -f ${rustserver}${rusthash} ] then touch ${rustserver}${rusthash} echo "$(date)" ${rustserver}rustserver monitor rm ${rustserver}${rusthash} fi ``` I made it wait for 180 seconds, before doing `rustserver monitor`, because in some "unfortunately timed" tests the server did not come back up fast enough and it restarted it again. The cronjobs I am running to monitor and restart the server: ``` * * * * * /home/rustserver/monitor-rustserver.sh >> /home/rustserver/log/uptime/monitor-rustserver.sh 2>&1 * * * * * sleep 10; /home/rustserver/monitor-rustserver.sh >> /home/rustserver/log/uptime/monitor-rustserver.log 2>&1 * * * * * sleep 20; /home/rustserver/monitor-rustserver.sh >> /home/rustserver/log/uptime/monitor-rustserver.log 2>&1 * * * * * sleep 30; /home/rustserver/monitor-rustserver.sh >> /home/rustserver/log/uptime/monitor-rustserver.log 2>&1 * * * * * sleep 40; /home/rustserver/monitor-rustserver.sh >> /home/rustserver/log/uptime/monitor-rustserver.log 2>&1 * * * * * sleep 50; /home/rustserver/monitor-rustserver.sh >> /home/rustserver/log/uptime/monitor-rustserver.log 2>&1 0 5 * * * wscat -c ws://SERVER_IP:PORT/PASSWORD -x '{"Identifier":1,Message:"global.restart 10",Name:"WebRcon"}' > /dev/null 2>&1 ``` If you want to use this... make sure you have the folders `/home/rustserver/.cache/` and `/home/rustserver/log/uptime/`, or modify the script and cronjobs to use whatever folders you want. Even though the `rusthash` should prevent the script to run `rustserver start` or `rustserver monitor` twice, I would not recommend running the monitoring script every second.