On Thu, Jan 17, 2008 at 10:39:33AM +0100, ADNET Ghislain wrote: > Hi, > > I have a little question about scheduling the vservers. I have a server: > > processor0 => Intel(R) Xeon(R) CPU 3040 @ 1.86GHz > processor1 => Intel(R) Xeon(R) CPU 3040 @ 1.86GHz > processorcount => 2 > productname => ProLiant DL320 G5 > > > and two vservers on it: from http://linux-vserver.org/ProcFS#sched > HOST:impec20:/usr/local/.aqadmin/home/aqadmin%(root)> more > /proc/virtual/40265/sched > FillRate: 6,0 > Interval: 10,10 > TokensMin: 15 > TokensMax: 125 > PrioBias: 0 > cpu 0: 9830 44 0 0 0 RI 62 15 125 6/10 0/10 0 0 > cpu 1: 6210 28 0 0 0 RI 62 15 125 6/10 0/10 0 0 ^----- ticks on hold > HOST:impec20:/usr/local/.aqadmin/home/aqadmin%(root)> more > /proc/virtual/40266/sched > FillRate: 4,0 > Interval: 10,10 > TokensMin: 15 > TokensMax: 125 > PrioBias: 0 > cpu 0: 3047 26 0 0 0 RI 62 15 125 4/10 0/10 0 0 > cpu 1: 7750 25 0 0 0 RI 62 15 125 4/10 0/10 0 0 ^----- ticks on hold > So basicaly i should have a 60/40 % for the servers, i disable idle cpu > by using a fillrate2 of 0. > > So now i run a cpuburn program on each vserver: > > top - 10:06:33 up 98 days, 23:15, 3 users, load average: 1.69, 0.73, 0.32 > Tasks: 113 total, 4 running, 109 sleeping, 0 stopped, 0 zombie > Cpu(s):100.0%us, 0.0%sy, 0.0%ni, 0.0%id, 0.0%wa, 0.0%hi, 0.0%si, > 0.0%st > Mem: 1033396k total, 956108k used, 77288k free, 220320k buffers > Swap: 2104496k total, 1032k used, 2103464k free, 574772k cached > > PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND > 16952 root 25 0 96 16 8 R 100 0.0 0:07.20 burnP6 > 16953 root 25 0 92 16 8 R 100 0.0 0:06.34 burnP6 > 16935 root 15 0 2232 1156 860 R 1 0.1 0:00.26 top > 1 root 15 0 1944 592 504 S 0 0.1 0:00.34 init > 2 root 10 -5 0 0 0 S 0 0.0 0:00.00 kthreadd > 3 root RT -5 0 0 0 S 0 0.0 0:00.14 migration/0 > 4 root 34 19 0 0 0 S 0 0.0 0:00.04 ksoftirqd/0 > > > Here i got a 50/50 distribution of the cpu. Is it normal ? Is it because > i have 2 cpu core and there is one more step to use for this case ? so it looks to me like you didn't enable hard scheduling for any of those two guests, which in turn, will give you roughly even distribution see http://linux-vserver.org/Capabilities_and_Flags HTH, Herbert > i use: > Versions: > Kernel: 2.6.22.9-vs2.2.0.4 > VS-API: 0x00020200 > util-vserver: 0.30.214; Oct 1 2007, 15:10:35 > > Features: > CC: gcc, gcc (GCC) 4.1.2 20061115 (prerelease) > (Debian 4.1.1-21) > CXX: g++, g++ (GCC) 4.1.2 20061115 (prerelease) > (Debian 4.1.1-21) > CPPFLAGS: '' > CFLAGS: '-g -O2 -std=c99 -Wall -pedantic -W > -funit-at-a-time' > CXXFLAGS: '-g -O2 -ansi -Wall -pedantic -W > -fmessage-length=0 -funit-at-a-time' > build/host: i686-pc-linux-gnu/i686-pc-linux-gnu > Use dietlibc: yes > Build C++ programs: yes > Build C99 programs: yes > Available APIs: v13,net,v21,v22,v23,netv2 > ext2fs Source: e2fsprogs > syscall(2) invocation: alternative > vserver(2) syscall#: 273/glibc > > > -- > Cordialement, > Ghislain