Re: [hlds_linux] TF2 ultra high cpu usage after the recent updates

2011-06-25 Thread Svensk Ljud Ljus Produktion
We have machines with both Debian and Fedora srcds-servers running and all is showing high spu usage. Its both TF2 and DoD:S running on them and we monitor all with munin. This is a pic from a debian 64-bit server with 4 cores running 2 dods servers.

Re: [hlds_linux] TF2 ultra high cpu usage after the recent updates

2011-06-25 Thread Marco Padovan
I found out my problem was within a kernel (running rt31 pachets) which kernel are you using? reverting to a non rt kernel solved the problem.. strange thing is: after the last update basically the servers always consumes the same amount of cpu when empty or when full... Il 24/06/2011 21:00,

Re: [hlds_linux] TF2 ultra high cpu usage after the recent updates

2011-06-25 Thread Marco Padovan
the problem are the preempt rt kernels... this is VERY problematic :| had to revert to a non preempt/rt kernel... but performance and game smoothes is bad now :( Il 25/06/2011 14:02, Svensk Ljud Ljus Produktion ha scritto: We have machines with both Debian and Fedora srcds-servers running and

Re: [hlds_linux] TF2 ultra high cpu usage after the recent updates

2011-06-25 Thread 1nsane
I'm not using RT kernels. On Sat, Jun 25, 2011 at 8:53 AM, Marco Padovan e...@evcz.tk wrote: the problem are the preempt rt kernels... this is VERY problematic :| had to revert to a non preempt/rt kernel... but performance and game smoothes is bad now :( Il 25/06/2011 14:02, Svensk Ljud

Re: [hlds_linux] TF2 ultra high cpu usage after the recent updates

2011-06-25 Thread Marco Padovan
How are your servers when empty? with rt preempt kernels they uses the same amount of cpu as when full (50% in my case) Il 25/06/2011 17:21, 1nsane ha scritto: I'm not using RT kernels. On Sat, Jun 25, 2011 at 8:53 AM, Marco Padovan e...@evcz.tk wrote: the problem are the preempt rt

Re: [hlds_linux] TF2 ultra high cpu usage after the recent updates

2011-06-25 Thread Marco Padovan
Tried on 2.6.33.7-rt29 and there it appear to be working fine... some compile flags... something's wrong somewhere :/ Il 25/06/2011 14:02, Svensk Ljud Ljus Produktion ha scritto: We have machines with both Debian and Fedora srcds-servers running and all is showing high spu usage. Its both

[hlds_linux] TF2 ultra high cpu usage after the recent updates

2011-06-24 Thread Marco Padovan
I'm getting impressively high cpu usage (50% cpu usage when server is empty) after the recents tf2 updates... these were just standard servers used for etf2l competitions (no mods, nothing... just STV enabled...) Anybody having the same problem? what's the cause?

Re: [hlds_linux] TF2 ultra high cpu usage after the recent updates

2011-06-24 Thread 1nsane
My servers (replay enabled) are maxing out the available CPU and experiencing some lag sometimes. These particular servers have a 2.83 ghz core each (Q9550) and prior to update worked fine. Absolutely no plugins are loaded yet. On Fri, Jun 24, 2011 at 10:50 AM, Marco Padovan e...@evcz.tk wrote:

Re: [hlds_linux] TF2 ultra high cpu usage after the recent updates

2011-06-24 Thread steve grout
ok, so first off we're on windows so appologies :) we have the same issue with high cpu usage, it also seems like the server has timed-out (the time out thing appears) when the server is doing the replay checks to see if it can delete. :( although we are running on the last MM and SM

Re: [hlds_linux] TF2 ultra high cpu usage after the recent updates

2011-06-24 Thread ics
Funny. I have the same CPU on our machines (replay enabled) and pre-update, there were spikes of lag but after this update, they are almost gone. No plugins currently on either. -ics 24.6.2011 21:16, 1nsane kirjoitti: My servers (replay enabled) are maxing out the available CPU and

Re: [hlds_linux] TF2 ultra high cpu usage after the recent updates

2011-06-24 Thread Arie
Experienced the same issue. Running 4 servers on a quad core box, server 1 is fine, takes 10% CPU idling, which is normal. The 3 others take between 30% and 50%. Doesn't matter if I shut them all down and just run them separately, the behaviour of the servers is the same. The first at 27015 will