On 01/09/17 21:15, Thomas C. Bishop wrote:
I messed up my raid5 array . I know a two of HDs are "failure
prediction" and one is out.. seagate is shipping me replacements.
This is my backup server so there's the actual copy of data but I'd
prefer to recover the array because other data scripts/tools have crept
into it ... mostly junk but would like to verify.
Here's out put as recommended at
https://raid.wiki.kernel.org/index.php/Linux_Raid
Thanks in advance for any assistance,
TOm
Okay, one failed drive, so it's not looking bad on that front. sdi1
seems to be the broken one.
Seagates - no mention of the model, or whether SCT/ERC is supported. Are
they Seagate NAS drives? Read the timeout mismatch / why you shouldn't
use desktop drives page on the wiki. Could that be the problem?
It's looking good in that nearly all the event counts are identical.
Seagate are sending a replacement? My immediate reaction is to wait
until it arrives, ddrescue sdi onto it, and then re-assemble the array.
It'll probably reject the new sdi because of the event mismatch, but it
might just work fine. If it does reject it, then you can do a --re-add
which because you've got a bitmap, should bring everything back hunky-dory.
Make sure ddrescue generates a log! If ddrescue can't copy the disk
completely, get back here with the contents of that log and we'll see if
we can mark the failed sectors as "bad" on the copy. That way, you can
safely re-add the drive knowing that a scrub will fall over the bad
sectors and re-create them correctly.
My gut feeling is that this should be a simple recovery, though if
you've actually got a spare drive on the array, you should have gone for
raid-6.
Cheers,
Wol
--
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
