[Top] [All Lists]

Recovering XFS data froma RAID failure

To: xfs@xxxxxxxxxxx
Subject: Recovering XFS data froma RAID failure
From: Mike Myers <mikesm559@xxxxxxxxx>
Date: Fri, 2 Jan 2009 10:36:34 -0800 (PST)
Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=X-YMail-OSG:Received:X-Mailer:Date:From:Subject:To:MIME-Version:Content-Type:Message-ID; b=jDsUi7kkg4ALemKWOeJEFxP0j2VkcbhLIfZuxg72JYx5UUmpy0fTz2vofRF5pmh2xdBFqlqYCE51giwLMUUV/KyENR2RjIOjdqMnQbNzcadl0gVEf1Df+97XUO7RfuEoIaBYzyFlXrdWLUMdfFQEOhZytOgXbF4mSTzJOtfhR10=;
Hello everyone. I seem to have run into a spate of bad luck on New Year's day 
morning, and would love some assistance.

I have (had) a 12 TB XFS filesystem running on top of a single LVM volume 
running on top of 2 7 TB software RAID5 sets (this is under opensuse 11.0).  
One of the RAID5 sets suffered a disk fault that had it go into degraded mode, 
and then suffered another disk fault while trying to restripe a new disk and a 
hardware SATA controller fault that occured all at the same time. Yes, I know 
the odds of this happening, and I am trying to discern what grevious sin I have 
committed to warrant this sort of treament.  I don't even work on Wall Street!

Anyways, I have a new SATA controller online and the disks are now talking to 
the system again, and I am working hard to try and salvage the array, but it 
looks pretty likely that RAID set may not be recoverable.  About 9 TB of data 
was present in the XFS volume, and the 2nd  7 TB RAID5 volume is working just 
fine.  I was wondering if there was any way to salvage data from the XFS volume 
so that any files who are intact on the half of the volume that is still there 
could be recovered?

I'm not trying to get back data that was rm'd, but if I can't recover one of 
the RAID5 volumes, it would be good to try and pull things off whatever is left.

Any ideas for me?



<Prev in Thread] Current Thread [Next in Thread>
  • Recovering XFS data froma RAID failure, Mike Myers <=