Help with start up script for Srcds

[zhelev81]

I need someone expirienced in linux to give me idea for a good start up script of the srcds server.

i need to do the following

run the server in gdb and when crashes to log the "bt" and restart the server again in gdb

I need this because my servers are crashing and hanging and i want to know what the cause is.

I will apreceate it if good script is given here to my needs

tnx in advance

[CannonessHannah]

I have some difficulty understanding what you're doing here, with the broken english. If you are using a shell script or other script to run SRCDS/HLDS, please show us the script so we can help diagnose it. We can't really help much unless you do.

Personally, I just run it in a seperate screen section since it will restart on crash. ie,
screen ./hlds_run -game gearbox -maxplayers 32 +map op4_repent -pingboost 2

[zhelev81]

I do not think my english is that bad,but i think you really have difficulty understanding the clear asked question

I did not say I'm using some kind of a script or whatever ....

I asked if someone can write one for me

run the server in a named screen in gdb

on a crash or hang to "bt" and log the output to a file,then restart the server in gdb again.

I hope is more clear to you now and i hope you can help me

[lol554]

I do not understand the question either, im just guessing. You want to run a server which puts logs in a file when it crashes? I would then put this in server.cfg

Code:

log on
sv_logfile 1(or logfile 1, cant clearly remember)

[CannonessHannah]

I think he wants a script to run the server .. except this already exists

Try ./srcds_run

If you want to make sure the server restarts across reboots, then add it to the list of startup applications.

[zhelev81]

omg i think I'm on a wrong forum lol

gdb is linux process crash back tracer .... this means you can see the last things what server did right before it crashed ... that is why i gave you link on my upper post and i think nobody knows what is it and you even did not have a look on what is it.

what is so dificult to undestand the explaned ?????

i need to run the server in gdb with different user and to restart the server when it crash because we all know that gdb pauses the process to let you analyse the crash ...

I think there is nothing to explane ,you just have to read it

[BehaartesEtwas]

zhelev81: there is an easier way of doing this. usually srcds (like any other process on linux) creates a core file when crashing. (if not it might be disabled by your distribution. either have a look at /etc/security/limits.conf or check with your distro how to change it) you can load this core file with gdb at any time after the crash (i.e. gdb srcds core.<some_number>). core files usually are created in the same directory as the srcds executable.

btw: don't shout at the others, they probably just don't know what gdb is

[zhelev81]

i do not shout but look how they answer me ... I had broken english and stuf like that ...

nobody wants this cores ,i know about them the people who want to help me are asking for gdb "bt" after the crash they are asking me to run my server in gdb and when it crashes to give them the "bt" output.

The reason i need this is that i do not have access to internet whole day and i do not want to loose my server to be down for 7-8 hours


thats why i need this auto gdb script so every time the server crashes to log the :"bt" to a file and then restart the server again in gdb ... people who knows linux and know how to do it to please help me with this

[BehaartesEtwas]

I perfectly understand what you want to do... you can run the "bt" after the crash if you load the core file with gdb. gdb will reconstruct the state of the process right after the crash when loading the core file, so it will behave very much the same as if you were inside gdb right after the crash. At least for the backtrace there will be no difference.