Yes, as far as serving hlds is concerned (and in my experience). Anyone running redhat owes it to themself to at least try the 2.4.9x kernel rpm. Worst case, they lose 20 minutes and a reboot and can go back to whatever kernel they were prviously using.
----- Original Message ----- From: "Oscar N aka 'Dreadful'" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Tuesday, March 18, 2003 11:54 AM Subject: Re: [hlds_linux] Linux Results - Performance Didn't someone discovered that the cpu usage changed very much in the kernel version above 2.4.9 or something like that? But there's nothing wrong with using 2.2.0 :) /Oscar Andy Hodges wrote: >Debian's TOP works fine. Even the user pings stayed normal with 20 users >compared to RH, etc. > >Give Debian a try with Kernel 2.2.20 .. don't take my word for it, find out >for yourself. > >-Andy Hodges > >----- Original Message ----- >From: "Jonas Andr�n" <[EMAIL PROTECTED]> >To: <[EMAIL PROTECTED]> >Sent: Tuesday, March 18, 2003 10:37 AM >Subject: Re: [hlds_linux] Linux Results - Performance > > > > >>sounds really strange that a 20player server under linux would only take >>10-15% cpu.... id rather think that there is something wrong with top in >>debian :) >> >>cs_assault is one of the heaviest maps for my server to run, while aztec >> >> >is > > >>one of the least heavy. >>which map are the procentages from? >> >>file://Jonas >>----- Original Message ----- >>From: "Andy Hodges" <[EMAIL PROTECTED]> >>To: <[EMAIL PROTECTED]> >>Sent: Tuesday, March 18, 2003 4:21 PM >>Subject: [hlds_linux] Linux Results - Performance >> >> >> >> >>>During the past week I have done some research and ran a few tests using >>>different platforms. Initially, I asked what Linux flavor everyone was >>> >>> >>using >> >> >>>and most people said RedHat, FreeBSD, and Debian. I installed all of >>> >>> >these > > >>>flavors on different linux machines. This testing was done on the >>>Counter-Strike MOD using 20 players. >>> >>>The maps were cs_assault, and de_aztec. >>> >>>All Linux flavors performed the same except for Debian. Debian's >>> >>> >>performance >> >> >>>was much better than the other machines. I used RedHat v7.2, RedHat v >>> >>> >7.3, > > >>>RedHat v8.0, Debian v3.0, and FreeBSD v4.7. >>> >>>20 player Counter-Strike Tests: >>> >>>RedHat v7.2, 7.3, and 8.0 = 80-90% CPU usage >>>FreeBSD = 60-70% CPU usage >>>Debian = 10-15% CPU usage >>> >>>The difference I found was the Kernel. RedHat installs the latest Kernel >>>which is v 2.4.18x. Debian installed v 2.2.20 as the default. I found >>> >>> >this > > >>>very interesting. So I installed Kernel 2.4.18x and the CPU usage went >>> >>> >up > > >>to >> >> >>>25%-35%. Still, beating RedHat and FreeBSD. >>> >>>I'm going to use Debian 3.0, and Kernel 2.2.20 (or 2.2.24) as my kernel. >>> >>> >>I'm >> >> >>>not saying that Debian is the best Linux out there, however, my tests >>> >>> >>prove >> >> >>>that Debian and Kernel 2.2.x make great harmony. >>> >>>Just thought I would pass the info. >>> >>>-Andy Hodges >>> >>> >>>_______________________________________________ >>>To unsubscribe, edit your list preferences, or view the list archives, >>> >>> >>please visit: >> >> >>>http://list.valvesoftware.com/mailman/listinfo/hlds_linux >>> >>> >>> >>_______________________________________________ >>To unsubscribe, edit your list preferences, or view the list archives, >> >> >please visit: > > >>http://list.valvesoftware.com/mailman/listinfo/hlds_linux >> >> > >_______________________________________________ >To unsubscribe, edit your list preferences, or view the list archives, please visit: >http://list.valvesoftware.com/mailman/listinfo/hlds_linux > > _______________________________________________ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux _______________________________________________ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
