https://bugzilla.novell.com/show_bug.cgi?id=821366 https://bugzilla.novell.com/show_bug.cgi?id=821366#c Ye Yuan changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |yyuan@suse.com AssignedTo|bnc-team-screening@forge.pr |nfbrown@suse.com |ovo.novell.com | -- 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.

https://bugzilla.novell.com/show_bug.cgi?id=821366 https://bugzilla.novell.com/show_bug.cgi?id=821366#c2 --- Comment #2 from Klaus Braun 2013-05-29 11:47:51 UTC --- For me it looks curious. On a test machine I tried to change something in the boot.md script like MDADM_DEVICE_TIMEOUT=480 Then after the next reboot I see different situations concerning mdmon: - one time I have a job like "mdmon md127" maybe coming from initrd ? - the other time I see @/sbin/mdmon -all --takeover --offroot coming from boot.md - but most times after a second and third reboot without any further change of boot.md mdmon isn't running at all. Furthermore I tried to create a softraid with mdadm --create. This works fine, even if mdon is not running after reboot. May be mdmon isn't necessary in this case, this is something I don't know. Creating an imsm based fake raid using mdadm --create --metadata=imsm I get the same problem as before: If mdmon isn't running the raid hangs. For me it looks like in most of the cases when I reboot my machine the raid isn't ready if systemd wants to start mdmon. In some rare cases mdmon could be started, but I don't see what should be done to reproduce this. -- 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.

https://bugzilla.novell.com/show_bug.cgi?id=821366 https://bugzilla.novell.com/show_bug.cgi?id=821366#c4 --- Comment #4 from Neil Brown 2013-05-30 22:00:08 UTC --- I expect that will be very helpful, thanks. I'll get back to you when I've had a chance to take a proper look. -- 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.

https://bugzilla.novell.com/show_bug.cgi?id=821366 https://bugzilla.novell.com/show_bug.cgi?id=821366#c6 Neil Brown changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO InfoProvider| |klaus.braun@geographie.uni- | |freiburg.de --- Comment #6 from Neil Brown 2013-06-03 22:16:11 UTC --- Would you be able to test the above RPM? - run mkinitrd again without "-f md", reboot and confirm that mdmon isn't running and the array is not accessable. - install the rpm rpm -U --force mdadm-3.2.6-0.x86_64.rpm - reboot and see if mdmon is running now. With this package mdmon is run as a systemd service rather than a simple forked process, so systemd knows not to kill it. Thanks. -- 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.

https://bugzilla.novell.com/show_bug.cgi?id=821366 https://bugzilla.novell.com/show_bug.cgi?id=821366#c8 Neil Brown changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED InfoProvider|klaus.braun@geographie.uni- | |freiburg.de | --- Comment #8 from Neil Brown 2013-06-05 21:06:12 UTC --- Thanks for testing - I'll schedule an update. I think that message at shutdown has always been there. It is because telling mdadmd to shut down md arrays deliberately doesn't - they just remain there until the final shutdown when the kernel dismantles them. This is intended. -- 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.

https://bugzilla.novell.com/show_bug.cgi?id=821366 https://bugzilla.novell.com/show_bug.cgi?id=821366#c10 Neil Brown changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO InfoProvider| |klaus.braun@geographie.uni- | |freiburg.de --- Comment #10 from Neil Brown 2013-06-11 00:22:57 UTC --- I think I was thinking of a different message - sorry. The "mdadmd" message should be completely independent of 'mdmon'. mdadmd.service runs "mdadm -F" which sends emails about failing devices and degraded arrays. It should be running whether or not mdmon is running, provided that "MDADM_SCAN" and "MDADM_MAIL" are set in /etc/sysconfig/mdadm If fsck is sometimes complaining and sometimes not then there must be some dependency problem, and systemd doesn't know to wait for something properly. I'll have to configure a machine with a similar setup and see. Just to confirm: You have '/' on a non-RAID device, and the Intel RAID array is separate from '/' - correct? -- 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.

https://bugzilla.novell.com/show_bug.cgi?id=821366 https://bugzilla.novell.com/show_bug.cgi?id=821366#c12 Benjamin Brunner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |RESOLVED InfoProvider|klaus.braun@geographie.uni- | |freiburg.de | Resolution| |FIXED --- Comment #12 from Benjamin Brunner 2013-06-24 08:37:03 CEST --- Update released for openSUSE 12.3. Resolved fixed. -- 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.