I'm running 3 x 24p servers inside a OpenVZ VPS which has 4 virtual
cores and have not done any fiddling of the scheduling. Servers run
perfectly fine with the defaults, constant sv_fps of 66.x all the time,
even when all servers are full.
Not sure what the underlying hardware is, /proc/cpuinfo reports:
model name : Intel(R) Core(TM) i7 CPU 960 @ 3.20GHz
On 3/11/2012 1:25 AM, Chris Oryschak wrote:
A recent thread made me wonder if i'm running my servers the most efficient
way.
For the longest time I've been running multiple servers on a single box
that has multiple cores. Each SRCDS instance would be assigned it's own
core to prevent it from hopping between cores as I assumed they might
overlap and max out that core. Plus I remember discussion of lag
spikes occurring when it moves between cores.
That being said, ICS in another thread said he's been running his servers
not assigned to any specific cores.
Here are my questions:
-Those of you who run multiple SRCDS instances per server, do you assign
them each to a core?
-Do you renice any processes?
-Do you change the realtime scheduling?
(
http://wiki.fragaholics.de/index.php/EN:Linux_Optimization_Guide#Setting_your_servers_to_run_with_realtime_scheduling)
-Any other process tweaks to give SRCDS more priority?
I'm curious how everyone else is doing this to achieve maximum performance.
Thanks.
Chris
_______________________________________________
To unsubscribe, edit your list preferences, or view the list archives, please
visit:
https://list.valvesoftware.com/cgi-bin/mailman/listinfo/hlds_linux
_______________________________________________
To unsubscribe, edit your list preferences, or view the list archives, please
visit:
https://list.valvesoftware.com/cgi-bin/mailman/listinfo/hlds_linux
