-----Original Message----- From: centos-bounces at centos.org [mailto:centos-bounces at centos.org] On Behalf Of Johnny Hughes Sent: Friday, May 13, 2005 1:04 PM To: CentOS ML Subject: RE: [CentOS] centos + kswapd0 error On Fri, 2005-05-13 at 12:45 -0600, mbeltran at americatel.com.sv wrote: > >> Hi all > >> > >> I have a box with Centos 4 and qmail with all pathches. since Last month > i > >> have had twice the same problem. The process kswapd0 uses all my > resources > >> of CPU and RAM memory and the consequence is that my local queue grow up > and > >> the box go slow too. > >> > >> The first time i guess the problem was the version of the kernel, so i > did > >> an upgrade form kernel.i686 2.6.9-5.0.3.EL to kernel.i686 2.6.9-5.0.5.EL > >> > >> apparently the problem was solved but today the problem came again. The > >> following snapshot shows the situation: > >> > >> > >> top - 10:53:48 up 16 days, 20:05, 1 user, load average: 119.27, 116.16, > >> 114.33 > >> Tasks: 528 total, 13 running, 515 sleeping, 0 stopped, 0 zombie > >> Cpu(s): 5.4% us, 89.4% sy, 0.0% ni, 0.3% id, 4.8% wa, 0.0% hi, 0.1% > si > >> Mem: 2075012k total, 2073404k used, 1608k free, 71056k buffers > >> Swap: 4192924k total, 4192924k used, 0k free, 36380k cached > >> > >> PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND > >> 60 root 17 0 0 0 0 D 75.4 0.0 585:22.79 kswapd0 > >> 2164 root 15 0 1736 484 432 S 31.0 0.0 31:57.07 > couriertcpd > >> 16334 vpopmail 19 0 2660 804 668 R 24.7 0.0 0:01.69 imapd > >> 2158 root 15 0 3264 488 432 S 23.4 0.0 1:17.51 > couriertcpd > >> 32608 apache 16 0 21604 8440 3016 R 15.5 0.4 1:38.09 httpd > >> 15284 vpopmail 18 0 4384 1032 760 R 14.9 0.0 0:08.32 > vdelivermail > >> > >> > >> > >> I really do not know what thing should i do for solve this problem > >> > >> thanks in advance > >> > >> Mario > > > >I think this bug is your problem: > > > >https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=150971 > > > >You can try the kernel at: > > > >http://people.redhat.com/davej/kernels/RHEL4/RPMS.kernel/ > > > >(or I can give you version 2.6.9-6.37.EL that is in 4Update1 beta) > > Hi Mr. Hughes > > The kernel version 2.6.9-6.37.EL has known issues >That is the only other one I have for EL4 (other than the released >kernels). >Maybe the 2.6.9-10.EL then ... but that is pretty new ... he just made >it yesterday :) Does anyone know if the following is related to the kswapd0 error outlined above? I have had my server freeze twice (under moderate load) and give the same error in my /var/log/messages. Or, if it is not related, does anyone have any ideas what might be causing my server to freeze (hard reboot needed to bring it back up). May 13 01:54:04 galactica kernel: Unable to handle kernel paging request at virtual address 00400000 May 13 01:54:04 galactica kernel: printing eip: May 13 01:54:04 galactica kernel: c013bb4b May 13 01:54:04 galactica kernel: *pde = 00000000 May 13 01:54:04 galactica kernel: Oops: 0000 [#1] May 13 01:54:04 galactica kernel: SMP May 13 01:54:04 galactica kernel: Modules linked in: parport_pc lp parport md5 i pv6 autofs4 sunrpc dm_mod button battery ac uhci_hcd hw_random e100 mii ext3 jbd May 13 01:54:04 galactica kernel: CPU: 0 May 13 01:54:04 galactica kernel: EIP: 0060:[<c013bb4b>] Not tainted VLI May 13 01:54:04 galactica kernel: EFLAGS: 00010097 (2.6.9-5.0.5.ELsmp) May 13 01:54:04 galactica kernel: EIP is at find_get_pages+0x30/0x50 May 13 01:54:04 galactica kernel: eax: 2002002c ebx: 00000004 ecx: 00000003 edx: 00400000 May 13 01:54:04 galactica kernel: esi: 0000000e edi: dedce638 ebp: c163ae94 esp: c163ae5c May 13 01:54:04 galactica kernel: ds: 007b es: 007b ss: 0068 May 13 01:54:04 galactica kernel: Process kswapd0 (pid: 31, threadinfo=c163a000 task=c1665670) May 13 01:54:04 galactica kernel: Stack: 00000000 c163ae8c 000000b4 00000000 c0143e10 c163ae94 dedce588 00000000 May 13 01:54:04 galactica kernel: c014423a 0000000e ffffffff dedce638 00000000 00000000 c1219e60 c13b8b80 May 13 02:13:47 galactica syslogd 1.4.1: restart.