I can't really comment on the rest, I have no experience with Windows servers. If its any aid to you, I run Debian 3.1 with (currently) the 2.6.12.3 kernel. I've always had a 2.6 kernel, even on Debian 3.0
Reading the below though, I don't believe your setup was truly that different from mine. At one point I DID run more than 3-4 game servers on the single machine, but that was back when a portion of them were usually empty. Looking at what you posted, you had on average 3 active game servers, with occasionally a 4th. The others were usually empty or private servers that idle most of the time. I should note that I moved my 4th game server to another machine as a proactive measure. The machine was not lagging ALL of the time, but it was occasionally getting hit very hard during primetime. It was an issue that was bad some nights and non-existant others, depending on the activity of the web side of things. All 4 servers are full of players. The only thing that truly matters in the end is how the machine behaves under full load. I can easily load up 20 servers but the machine will die if they're actually used. I'm just throwing this out as well, I can't say this is truth: Just because the server uses "5%" on Windows and "30%" on Linux, does that really mean the OS is performing poorly? Could it also mean that Linux is simply allowing the process to use more CPU/frames because its available? That's probably why you asked for people to post "stats" output though. > I was running this. > > 1 - 20 Man cs 1.6 server - fairly full most of the time. > 3 - 16 Man cs 1.6 servers - 2 - full alot of the time - 1 - not hardly > full at all > 1 - 20 Man Source server - never full > 1 - 24 Man dod server - most of the time not full. but occassionally we > would get full 24 people on > 1 - 12 man private 1.6 server _______________________________________________ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
