Mailinglist Archive: opensuse-bugs (5243 mails)

< Previous Next >
[Bug 793954] raid1 array sometimes started with one of two disks at boot (again)

https://bugzilla.novell.com/show_bug.cgi?id=793954

https://bugzilla.novell.com/show_bug.cgi?id=793954#c21


Neil Brown <nfbrown@xxxxxxxx> changed:

What |Removed |Added
----------------------------------------------------------------------------
Status|ASSIGNED |NEEDINFO
InfoProvider| |suse-beta@xxxxxxxxx

--- Comment #21 from Neil Brown <nfbrown@xxxxxxxx> 2013-03-13 01:03:20 UTC ---
Actually I don't have a completely clear understanding.

I can see the races and the need to avoid them... but they still shouldn't be
happening.

The race can only be between "mdadm -I" run by udev, and "mdadm -As" run by
boot.md.
However boot.md call 'udevadm settle' which should wait for all udev activity
to complete. So "mdadm -I" shouldn't run concurrently to "mdadm -As".

I've found a case where "udevadm settle" won't actually wait for the queue to
settle, but it seems an unlikely case. It is worth checking though.

So: could you please edit your "/etc/init.d/boot.md" and replace the line

/sbin/udevadm settle --timeout="$MDADM_DEVICE_TIMEOUT"

with

/sbin/udevadm --debug settle --timeout="$MDADM_DEVICE_TIMEOUT" >
/dev/kmsg 2>&1

This should cause "calling: settle" to appear in the "dmesg" output.
If "timeout waiting for udev queue" appears as well then that points to the
problem. However whatever appears there might be useful.

--
Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

< Previous Next >