Michel,
At 10:27 Uhr -0500 18.11.1999, Michel Pelletier wrote:
>Ok, there are three groups of messages for each of the three md devices.
>Here is the group for the failed device:
>
>...
>
>Nov 16 13:36:28 korak kernel: running: <sdb4><sda4>
>Nov 16 13:36:28 korak kernel: now!
>Nov 16 13:36:28 korak kernel: sdb4's event counter: 00000016
>Nov 16 13:36:28 korak kernel: sda4's event counter: 00000017
>Nov 16 13:36:28 korak kernel: md: superblock update time inconsistency
>-- using
>the most recent one
>Nov 16 13:36:28 korak kernel: freshest: sda4
>Nov 16 13:36:28 korak kernel: md2: kicking faulty sdb4!
>Nov 16 13:36:28 korak kernel: unbind<sdb4,1>
>Nov 16 13:36:28 korak kernel: export_rdev(sdb4)
>Nov 16 13:36:28 korak kernel: raid1: raid set md2 active with 1 out of 2
>mirrors
...
Beware - I am using these things for only two days now and have my
own problems, too...
but this seems like something I know...
if you cat /proc/mdstat it will show you that /dev/md2 is running
with only one of the two disks.
For any reason, it was not in the raid the session before and data is
out of sync.
I had this situation and did not know why one of my disks was not
part of the raid. But a
raidhotadd /dev/md2 /dev/sdb4 should insert it back into the raid and
update its content correctly.
I would say that's what the event counters are for...
You have to make sure your /etc/raidtab is correctly set up!
Especially /dev/sdb4 may not be marked
as failed-disk.
cat /proc/mdstat should tell you that your raid is beeing
reconstructed after the raidhotadd
Jochen
