Re: Wiki-recovering failed raid, overlay problem

[Phil Turmel <philip@turmel.org>]

On 06/01/2013 08:40 PM, Chris Finley wrote:
> On Sat, Jun 1, 2013 at 4:30 PM, Phil Turmel <philip@turmel.org> wrote:
>> Hi Chris,
>>
>> On 06/01/2013 02:23 AM, Chris Finley wrote:
>>> I am trying to recover a failed Raid 5 array by following the guide at
>>> https://raid.wiki.kernel.org/index.php/Recovering_a_failed_software_RAID
>>
>> Stop.  Report the *critical* details of your setup.  At least:
> 
> Thank you for the reply.
> 
> Oh, yes. I'm the guy from an earlier post:
> http://marc.info/?l=linux-raid&m=136840333618808&w=2

I missed it--I must have been busy.

> Because each of the drives had some read errors, I thought it would be
> safer to make the first attempt with overlays. There is always the
> possibility of me entering command incorrectly too :)

As long as the original metadata is still present, mdadm is quite
robust.  Overlays are useful when you don't know the original metadata
properties and don't have enough spare drives.

The material provided is quite complete, but lacks a correlation between
device names and drive serial numbers.  I'd like some more confidence there:

Please show the output of my 'lsdrv' script [1] as your system is now
set up.

Your drive with S/N S2H7JD2B105688 seems to be the worst, with
triple-digit pending sectors.  This suggests a mismatch between your
drives' error correction time limits and the linux drivers' default
timeout.  And a lack of regular scrubbing to clean up pending sectors.
"smartctl -l scterc" for each drive would give useful information.
Anyways, the drive may not be really failing--it has zero relocations.

If S2H7JD2B105688 was the old /dev/sdd, then it doesn't matter, but
you've now lost the opportunity to correct those sectors.

Phil

[1] http://github.com/pturmel/lsdrv/