[Wachu]

Hi,
Since 2 days, my TF2 server started crashing with this message:

RELIABLE SNAPSHOT OVERFLOW! Triggering trace to see what is so large ./srcds_run: line 372: 31368 File size limit exceeded$HL_CMD

This crashing is random. Server changes map few times and gets crashed.
Any idea how can I fix this?
I have 2 similar TF2 servers. Both have the same error.
Error log's from SM are clear.
SM version: 1.4.4 stable
MM: 1.8.7 Stable
Greetings from Poland.
Wachu

[PreFix]

Same for us.

[pheadxdll]

I got this exact same crash, check out your tf/ directory and look for a file called netspike.txt. Mine was up to the filesize limit of 2GB. I deleted it and haven't crashed so far. Give it a try. I get this backtrace for one of my coredumps:

Program terminated with signal 25, File size limit exceeded.
#0  0xffffe410 in __kernel_vsyscall ()
(gdb) bt
#0  0xffffe410 in __kernel_vsyscall ()
#1  0x008eb09b in write () from /lib/libc.so.6
#2  0x0088ae34 in _IO_new_file_write () from /lib/libc.so.6
#3  0x0088aaf5 in new_do_write () from /lib/libc.so.6
#4  0x0088adaf in _IO_new_file_xsputn () from /lib/libc.so.6
#5  0x0088128f in fwrite () from /lib/libc.so.6
#6  0xf799018b in ?? ()
#7  0x12488b60 in ?? ()
#8  0x00000001 in ?? ()
#9  0x00011978 in ?? ()
#10 0x0980aa80 in ?? ()
#11 0x0d000000 in ?? ()
#12 0x00000000 in ?? ()

It was trying to write to the file apparently. In the mailing list fletcher says it supposed to debug entities that caused players to be dropped from a snapshot overflow but others are saying that it starts to log every packet. I would create a cron job to remove it before it gets too large.

[PreFix]

Anyone can write manual how to make that cron job?
Anyway thanks , pheadxdll , for sharing that good information