What filesystem is on /dev/md1 (the sd*5 disks)?
I observe similar behaviour (large IO, none through the md device) on my
server running 10.04. So I don't believe you should be seeing the IO go
via md* devices in iotop:-
avg-cpu: %user %nice %system %iowait %steal %idle
0.49 0.00 4.39 95.12 0.00 0.00
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn
sda 64.00 0.00 520.00 0 1040
sdb 78.50 0.00 664.00 0 1328
sdc 64.50 0.00 536.00 0 1072
sdd 67.50 0.00 560.00 0 1120
sde 0.00 0.00 0.00 0 0
md0 0.00 0.00 0.00 0 0
dm-0 0.00 0.00 0.00 0 0
md0 : active raid10 sda1[2] sdc1[3] sdd1[0] sdb1[1]
3907023872 blocks 64K chunks 2 near-copies [4/4] [UUUU]
Periodically I do see this:-
avg-cpu: %user %nice %system %iowait %steal %idle
0.50 0.00 3.75 95.75 0.00 0.00
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn
sda 42.50 0.00 1016.00 0 2032
sdb 75.50 0.00 716.00 0 1432
sdc 22.50 0.00 856.00 0 1712
sdd 39.00 0.00 800.00 0 1600
sde 0.00 0.00 0.00 0 0
md0 7733.50 0.00 61868.00 0 123736
dm-0 7674.50 0.00 61396.00 0 122792
(high md0 tps) but only for brief periods.
Sorry if this isnt helpful, just trying to see if you're barking up the
wrong tree with the md0 io post at the bottom.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/815540
Title:
Server becomes unresponsive after spawning 16 ksoftirqd processes
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/815540/+subscriptions
--
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
