Ok, i could "work around" the problem in my case, by just disabling the
"change" ACTION in
/lib/udev/rules.d/95-multipath.rules
as discussed in
https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/571093/comments/9
In my case, two FiberChannel adapters with access to a SAN device over 4
different paths which are then handled by multipathd, udev was constantly
"discovering" the 4 multipath devices (sdb/sdc/sdd/sde) and causing a high CPU
load of around 3 (8 CPU system). With "ps" i could see a process tree like this
again and again:
393 ? S<s 22:09 udevd --daemon
20696 ? S< 0:12 \_ udevd --daemon
7171 ? S< 0:00 | \_ /sbin/multipath -v0 /dev/sdb
7190 ? R< 0:00 | \_ /sbin/mpath_prio_emc /dev/sdd
20702 ? S< 0:10 \_ udevd --daemon
7177 ? S< 0:00 | \_ /sbin/multipath -v0 /dev/sdc
7189 ? R< 0:00 | \_ /lib/udev/scsi_id -g -u -d /dev/sdb
31869 ? S< 0:01 \_ udevd --daemon
7179 ? S< 0:00 | \_ /sbin/multipath -v0 /dev/sdd
7191 ? R< 0:00 | \_ /sbin/mpath_prio_emc /dev/sdb
31875 ? S< 0:01 \_ udevd --daemon
After having disabled the "change" ACTION in
/lib/udev/rules.d/95-multipath.rules
the number of events got reduced from "tons per second" to one every 30 seconds
in my hardware.
Thus I conclude that all the other events were spurious events generated by a
udev/multipath feedback loop...
However, there is still some additional issue in the kernel/QLA-driver probably:
the same hardware setup under Karmic doesn't trigger _any_ event in udev,
not even that one every 30 seconds!
--
Ubuntu 10.04 32bit. Udev high cpu usage when it is started also on AMD64 debian
testing
https://bugs.launchpad.net/bugs/578180
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
--
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
