Bug 818501 - md raid occasionally incomplete at startup
Summary: md raid occasionally incomplete at startup
Status: RESOLVED DUPLICATE of bug 793954
Alias: None
Product: openSUSE 12.3
Classification: openSUSE
Component: Basesystem (show other bugs)
Version: Final
Hardware: i686 openSUSE 12.3
: P5 - None : Normal (vote)
Target Milestone: ---
Assignee: E-mail List
QA Contact: E-mail List
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2013-05-04 17:31 UTC by Forgotten User 5bOUleMVRM
Modified: 2013-05-12 15:41 UTC (History)
1 user (show)

See Also:
Found By: ---
Services Priority:
Business Priority:
Blocker: ---
Marketing QA Status: ---
IT Deployment: ---


Attachments
/proc/mdstat (608 bytes, text/plain)
2013-05-04 17:31 UTC, Forgotten User 5bOUleMVRM
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Forgotten User 5bOUleMVRM 2013-05-04 17:31:38 UTC
Created attachment 537943 [details]
/proc/mdstat

User-Agent:       Mozilla/5.0 (X11; Linux i686; rv:20.0) Gecko/20100101 Firefox/20.0

md raid devices sometimes get assembled incompletely at startup for unknown reason. This applies to both desktop computers of mine since I have installed openSUSE12.3. Even the latest updates from May 3th did not solve this problem. I have been using these devices for years with openSUSE, but before 12.3 this fault has never occured.

When the fault occurs, the missing partition can be added afterwards to the degraded array by "mdadm --add /dev/md1 /dev/sda3" without problems.

I suspect that systemd tries to acticate the raid devices before all disk partitions have been scanned, but have no idea how to fix it. If you provide instructions, I will be glad to do further testing.

/proc/mdstat:

Personalities : [raid1] [raid0] [raid10] [raid6] [raid5] [raid4] 
md3 : active (auto-read-only) raid1 sda6[0] sdb6[1]
      103412 blocks super 1.0 [2/2] [UU]
      bitmap: 0/1 pages [0KB], 65536KB chunk

md0 : active raid1 sda7[0] sdb7[1]
      89617336 blocks super 1.0 [2/2] [UU]
      bitmap: 0/1 pages [0KB], 65536KB chunk

md1 : active (auto-read-only) raid1 sdb3[1]
      20482800 blocks super 1.0 [2/1] [_U]
      bitmap: 0/157 pages [0KB], 64KB chunk

md2 : active raid1 sdb2[1] sda2[0]
      20482800 blocks super 1.0 [2/2] [UU]
      bitmap: 1/1 pages [4KB], 65536KB chunk

Output of systemctl command related to md devices:

sys-devi...ock-md0.device loaded active plugged       /sys/devices/virtual/block/md0
sys-devi...ock-md1.device loaded active plugged       /sys/devices/virtual/block/md1
sys-devi...ock-md2.device loaded active plugged       /sys/devices/virtual/block/md2
sys-devi...ock-md3.device loaded active plugged       /sys/devices/virtual/block/md3

/etc/mdadm.conf:

DEVICE /dev/sda* /dev/sdb*
ARRAY /dev/md1 UUID=10c839d1:c7c910ca:76be2406:eb0790d3
ARRAY /dev/md2 UUID=228828a6:0f4beb07:d5445588:bb63f35b
ARRAY /dev/md0 UUID=7bb819f5:e827b8ae:40a32fe0:e9a6cf62
ARRAY /dev/md3 UUID=de1b1ac0:6ab6ff46:9d569c60:76bc7dcf


Reproducible: Sometimes

Steps to Reproduce:
1.
2.
3.
Comment 1 Christian Boltz 2013-05-11 18:32:13 UTC
Sounds like an issue I know very well...

*** This bug has been marked as a duplicate of bug 793954 ***
Comment 2 Forgotten User 5bOUleMVRM 2013-05-12 15:41:54 UTC
I haven't noticed the other bug report because I searched only for bugs related to openSUSE 12.3, but indeed, it looks very similar.

I just applied the changes from comment 30, but unfortunately, the fault occurred quite seldom (perhaps 1 in 100 times) and many starts are required to be confident that it is gone.

Thanks for the tip!