RAID array broken on boot

dannewestis · 01-18-2010, 10:14 AM

Hi,

I have an issue with a RAID array failing on boot. Seems like an issue with the file system.

I get passed the RAID BIOS (and from what I can see, it looks alright there, all devices appear), but then the following error messages appear:

Code:

raid5: failed to run raid set md0
mdadm: failed to RUN_ARRAY /dev/md0 input/output error
mdadm: Not enough devices to start the array

and further down:

Code:

fsck.ext3: Invalid argument when trying to open /dev/md0
/dev/md0:
The superblock could not be read or does not describe a correct ext2
filesystem.  If the device is valid and it really contains an ext2
filesystem (and not swap or ufs or something else), then the superblock
is corrupt, and you might try running e2fsck with an alternate superblock:
    e2fsck -b 8193 <device>

I then login with root password to get a "Repair file system" prompt. Tried with fsck, but not working.

It's 4x1TB in RAID5 on HighPoint RocketRAID 2300 4P SATA II/300, with Fedora 9. Not sure what other system info might be needed.

Would be really grateful to how to solve this issue, as we have lots of important files on that RAID...

Thanks!

/Daniel

r0b0 · 01-19-2010, 03:35 AM

Can you post contents of /proc/mdstat ? From the first message it appears that more than one drive is marked as failed.

dannewestis · 01-19-2010, 03:44 AM

Quote:

Originally Posted by r0b0

Can you post contents of /proc/mdstat ? From the first message it appears that more than one drive is marked as failed.

Thanks for your reply.

Here's what cat /proc/mdstat gives:
Personalitieis : [raid6] [raid5] [raid4]
md0 : inactive sdb1[0] sdd1[2]
1953519616 blocks

mdadm --examine gives this:
Number Major Minor RaidDevice State
0 8 17 0 active synd /dev/sdb1
1 0 0 1 faulty removed
2 8 49 2 active sync /dev/sdd1
3 0 0 3 faulty removed

Seems like 2 out of 4 are working, which is 1 too less for RAID5... But can I save what's on the RAID still?

dannewestis · 01-19-2010, 07:56 AM

Anyone's got a clue on this one? Are my data irreparable? Since fdisk finds all four, I suppose it should be a software issue? But I'm not sure what to do, to save what's on there.

r0b0 · 01-21-2010, 03:30 AM

If you are certain that the drives are not faulty, you can try marking them as clean. I think the command is something like mdadm manage --clean /dev/sda1 --clean /dev/sdc1

dannewestis · 01-21-2010, 02:45 PM

Quote:

Originally Posted by r0b0

If you are certain that the drives are not faulty, you can try marking them as clean. I think the command is something like mdadm manage --clean /dev/sda1 --clean /dev/sdc1

Still doesn't work... I can create the array /dev/md0. But when running

Code:

fsck.ext3 -v /dev/md0

I get "Group descriptors look bad... trying backup blocks". And then lots and lots of questions about relocating blocks, inod table for group xxx is not in the group etc. that I have to answer yes to.

I suppose there's no hope? I mean, I am not SURE all disks are ok, only that the raid BIOS and fdisk finds them. Should I better give up or are there any options left?

I've realized that most of the files are backed up or possible to recreate, although there still are quite a few important files that I very much would like to recover.