On Wed, 12 Dec 2012, Bernd Waage wrote:
I run an 8-disk raid 6 on which sporadically 2 drives dropped out, that
I could just re-add when I zeroed the zuperblock beforehand. Recently,
upon re-adding those 2 drives (after the zero-superblock) a third drive
dropped out after 5-10 minutes of syncing. I then did a zero-superblock
on the third drive and tried to re-add it - which failed.
At this point, your only option is to re-create the raid superblocks via
the --create command. Since you took your RAID6 (which can handle 2 drives
going bad) with 2 bad drives, and then zero:ed the superblock on a third
drive, you now have a non-working RAID6.
What did you expect to happen when you zero:ed the third drive?
I would imagine that the third drive dropped out due to read error? Why
did the other two drives drop out?
Anyhow, you now have two options. You can re-create the raid6 with two
missing drives, meaning you still have the read error somewhere (if that's
what happened), and try to work around that (perhaps dd_rescue the drive
to a non-defective drive, losing the information on the bad blocks), or
you can decide that very little has changed since the first two drives
dropped out, and use one of those instead of the drive with the read
errors.
Please read up carefully regarding data mdadm versions differing with
default data offset, chunk size etc. Recreating with --create
--assume-clean is a big dangerous hammer that might wipe all your data.
Mess up the order of the drives and forget --assume-clean or write to the
array in this state, and you might destroy things.
If you value your data, be patient and make sure you understand all the
ramifications of what you're doing. It might be a good thing to back up
all drives before continuing (if you have the economic possibility of
doing so).
--
Mikael Abrahamsson email: swmike@xxxxxxxxx
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
