FAQ Search Today's Posts Mark Forums Read
» Video Reviews

» Linux Archive

Linux-archive is a website aiming to archive linux email lists and to make them easily accessible for linux users/developers.


» Sponsor

» Partners

» Sponsor

Go Back   Linux Archive > Debian > Debian User

 
 
LinkBack Thread Tools
 
Old 04-18-2012, 10:07 AM
Ramon Hofer
 
Default RAID5 failure; possible to repair?

Hi all

I have a home server with a raid5.
After switching to a new case my graphics card died and I replaced it
with an old but very powerful card just to try things.

Unfortunately the PSU wasn't strong enough and there were some crashes.

I could fix the problems of the other disks but I had no luck with the
raid.

What I did is:

First I checked with `mdadm --detail /dev/md0` and everything seemed fine
(raid clean).
But it wasn't possible to mount md0 so I ran `fsck /dev/md0` which was
busy for quite a while. It removed one disk after the other from the
raid. Afterwards I ended with an empty raid with four spare disks and now
I don't even have a md0 device.

Is there anything I can do to recover the files from there or should I
start from scratch?


Best regards
Ramon


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: jmm3pt$33b$1@dough.gmane.org">http://lists.debian.org/jmm3pt$33b$1@dough.gmane.org
 
Old 04-18-2012, 01:03 PM
Daniel Koch
 
Default RAID5 failure; possible to repair?

Zero all the superblocks on all the disks. (mdadm --zero-superblock /dev/sd{b..d})Recreate the array with the "--assume-clean" option. (mdadm --create --verbose /dev/md0 --auto=yes --assume-clean --level=5 --raid-devices=3 /dev/sdb /dev/sdc /dev/sdd)
Mark it possibly dirty with: "mdadm --assemble /dev/md0 --update=resync"Let it resync
Mount it and see if it is restored
If this is not working your raid is lost.
 
Old 04-18-2012, 02:52 PM
Ramon Hofer
 
Default RAID5 failure; possible to repair?

On Wed, 18 Apr 2012 15:03:04 +0200, Daniel Koch wrote:

> - Zero all the superblocks on all the disks. (mdadm --zero-superblock
> /dev/sd{b..d})
> - Recreate the array with the "--assume-clean" option. (mdadm
> --create --verbose /dev/md0 --auto=yes --assume-clean --level=5
> --raid-devices=3 /dev/sdb /dev/sdc /dev/sdd)
> - Mark it possibly dirty with: "mdadm --assemble /dev/md0
> --update=resync"
> - Let it resync
> - Mount it and see if it is restored

Thanks alot for your answer, Daniel.
I had four devices in the raid /dev/sd{b..e}1. So I did the this:

http://pastebin.com/raw.php?i=8BJXMqTV

I fear when I ran fsck yesterday it deleted the raid or the filesystem.
It told me that it had some errors and will continue with three devices.
Still in the progress it told me it will continue with two, then with one
and in the end with zero. Then I ended up with the raid I mentioned in my
first post with no discs but four spares.

fdisk tells me that md0 doesn't contain a valid partition table.
Can I try to repair the table?


Best regards
Ramon


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: jmmkeh$c2q$1@dough.gmane.org">http://lists.debian.org/jmmkeh$c2q$1@dough.gmane.org
 

Thread Tools




All times are GMT. The time now is 04:12 PM.

VBulletin, Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.
Copyright 2007 - 2008, www.linux-archive.org