OpenSUSE 11.1: Dovecot IMAP periodically goes into state D

English Applications
1

Hi All,

It has been a month I’m trying to figure out what is going on with my opensuse server. About once a month dovecot’s imap daemon is going into state D (uninteruptable sleep if I’m not mistaken) taking up server resources. Normally several instances “die” this way so at the end server becomes irresponsive since CPU load goes up to 40 or higher. It’s impossible to kill these processes. The only solution is to reboot (if it’s not too late) or power off-power up.

Server logs do not show anything unusual, so do not see what could lead into this situation.

Anyone have seen this issue or have a clue where to search?

Thanks in advance!

2

happened again today. Except this time imap haven’t gone into state D, but still when I tried to kill it the server stopped responding.

Here is my ‘top’ log:

06:04:05 everthing yet normal:

top - 06:04:05 up 4 days, 20:51, 0 users, load average: 0.00, 0.00, 0.00
Tasks: 99 total, 1 running, 98 sleeping, 0 stopped, 0 zombie
Cpu(s): 0.1%us, 0.0%sy, 0.0%ni, 99.7%id, 0.2%wa, 0.0%hi, 0.0%si, 0.0%st
Mem: 4043472k total, 3814576k used, 228896k free, 328764k buffers
Swap: 2104472k total, 64k used, 2104408k free, 3238324k cached

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
7 root 15 -5 0 0 0 S 0 0.0 1:47.47 events/0
3245 mysql 20 0 211m 26m 5252 S 0 0.7 2:54.89 mysqld
2688 root 20 0 8908 500 372 S 0 0.0 1:08.01 irqbalance
7476 postfix 20 0 98.7m 4568 3300 S 0 0.1 0:00.04 smtpd
1 root 20 0 1064 384 324 S 0 0.0 0:02.90 init
2 root 15 -5 0 0 0 S 0 0.0 0:00.00 kthreadd
3 root RT -5 0 0 0 S 0 …

06:05:45 - here comes the trouble:
top - 06:05:45 up 4 days, 20:53, 0 users, load average: 0.93, 0.24, 0.08
Tasks: 96 total, 3 running, 93 sleeping, 0 stopped, 0 zombie
Cpu(s): 0.5%us, 16.3%sy, 0.0%ni, 82.3%id, 0.9%wa, 0.0%hi, 0.0%si, 0.0%st
Mem: 4043472k total, 3804164k used, 239308k free, 328860k buffers
Swap: 2104472k total, 64k used, 2104408k free, 3238360k cached

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
3967 postfix 20 0 6936 1500 1076 R 32 0.0 0:33.49 imap
3245 mysql 20 0 211m 26m 5252 S 0 0.7 2:54.93 mysqld
7 root 15 -5 0 0 0 S 0 0.0 1:47.50 events/0
8 root 15 -5 0 0 0 R 0 0.0 0:13.16 events/1
21 root 20 0 0 0 0 S 0 0.0 0:04.66 pdflush
2688 root 20 0 8908 500 372 S 0 0.0 1:08.03 irqbalance
2744 wwwrun 20 0 275m 13m 3288 S 0 0.3 0:00.48 httpd2-prefork
3297 root 20 0 269m 10m 5792 S 0 0.3 0:05.86 httpd2-prefork
1 root 20 0 1064 384 324 S 0 0.0 0:02.90 init
…

proc 3967 starts to consume cpu…
and it goes up to ~ 14 at 9 am:

top - 09:22:06 up 5 days, 9 min, 2 users, load average: 14.82, 7.03, 3.98
Tasks: 106 total, 5 running, 101 sleeping, 0 stopped, 0 zombie
Cpu(s): 0.0%us, 50.0%sy, 0.0%ni, 6.7%id, 43.2%wa, 0.0%hi, 0.0%si, 0.0%st
Mem: 4043472k total, 3829076k used, 214396k free, 330344k buffers
Swap: 2104472k total, 64k used, 2104408k free, 3242128k cached

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
3967 postfix 20 0 6936 1500 1076 R 100 0.0 195:13.21 imap
3245 mysql 20 0 211m 26m 5256 S 0 0.7 2:58.92 mysqld
1846 root 20 0 19248 1228 772 D 0 0.0 0:20.88 syslog-ng
2688 root 20 0 8908 500 372 S 0 0.0 1:09.91 irqbalance
1 root 20 0 1064 384 324 S 0 0.0 0:02.94 init

3

Can you please post the result of:

uname -a

to show which kernel you have.

4

Have you applied all the kernel updates, in particular, have you updated the kernel to anything more recent than about March 2009? There is a known issue with the kernel on the distro media.

5

sorry indeed forgoten uname:

Linux myserver 2.6.27.7-9-default #1 SMP 2008-12-04 18:10:04 +0100 x86_64 x86_64 x86_64 GNU/Linux

I’ve tried to apply kernel patch upon standart 11.1 but it failed… I had problem with boot so reverted to standart kernel. Perhaps time to move onto 11.2

6

2.6.27.7-9

This kernel is too old, contains the bug.

7

Just upgraded to
Linux myserver 2.6.27.39-0.2-default #1 SMP 2009-11-23 12:57:38 +0100 x86_64 x86_64 x86_64 GNU/Linux

Will see if this stops freaking freezes of the server.