How to recover from massive disk failure

How to recover from massive disk failure

[Jacob Schmidt Madsen]

Hey

I have a 2 controllers: Promise SATA150 TX4 and Promise SATA300 TX4.

2 disks are connected to the SATA150 and 4 disks are connected to SATA300.

The 6 disks are part of a raid5 array.

Today the SATA300 controller failed and the 4 disks were excluded from the 
array.

The array was inactive the moment the 4 disks were excluded. After rebooting 
all controllers and disks are online again.

But now i have trouble starting the array and i could really use some input 
from smarter minds on this list.

Here's some output i gathered:

# cat /proc/mdstat:
md5 : active raid5 sdg1[6](F) sdf1[7](F) sde1[8](F) sdd1[9](F) sdc1[5] sdb1[4]
1562842880 blocks level 5, 64k chunk, algorithm 2 [6/2] [____UU]

The 4 failed disks are the ones connected to the failed controller.

The following is after the reboot.

# cat /proc/mdstat:
md5 : inactive sdb1[4] sdc1[5] sdg1[3] sdf1[2] sde1[1] sdd1[0]
1875411456 blocks

I then did the following in hope that it would help:

# mdadm -S /dev/md5
mdadm: stopped /dev/md5
# mdadm -As /dev/md5
mdadm: /dev/md5 assembled from 2 drives - not enough to start the array.

No luck!

What can i do to get it up and running again?

Thanks!

Re: How to recover from massive disk failure

[Neil Brown]

On Friday November 10, jacob@mungo.dk wrote:
> # mdadm -As /dev/md5
> mdadm: /dev/md5 assembled from 2 drives - not enough to start the array.
> 
> No luck!
> 
> What can i do to get it up and running again?

Add a "--force" flag.

NeilBrown

Re: How to recover from massive disk failure

[Jacob Schmidt Madsen]

Thank you very much for the quick answer!

One last one...

This is how my entry in mdadm.conf look:
DEVICE /dev/sd[bcdefg]1
ARRAY /dev/md5 level=raid5 num-devices=6 
UUID=a4a5dae9:04a09c3a:cd3fd7be:b754f4fe 
devices=/dev/sdb1,/dev/sdc1,/dev/sdd1,/dev/sde1,/dev/sdf1,/dev/sdg1

Do the order of the devices matter?

Thanks again!

On Friday 10 November 2006 01:54, Neil Brown wrote:
> On Friday November 10, jacob@mungo.dk wrote:
> > # mdadm -As /dev/md5
> > mdadm: /dev/md5 assembled from 2 drives - not enough to start the array.
> >
> > No luck!
> >
> > What can i do to get it up and running again?
>
> Add a "--force" flag.
>
> NeilBrown
> -
> To unsubscribe from this list: send the line "unsubscribe linux-raid" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: How to recover from massive disk failure

[Neil Brown]

On Friday November 10, jacob@mungo.dk wrote:
> Thank you very much for the quick answer!
> 
> One last one...
> 
> This is how my entry in mdadm.conf look:
> DEVICE /dev/sd[bcdefg]1
> ARRAY /dev/md5 level=raid5 num-devices=6 
> UUID=a4a5dae9:04a09c3a:cd3fd7be:b754f4fe 
> devices=/dev/sdb1,/dev/sdc1,/dev/sdd1,/dev/sde1,/dev/sdf1,/dev/sdg1
> 
> Do the order of the devices matter?

No, but it is best not to have them.
i.e. get rid of the "devices=" bit and change the DEVICE line to
   DEVICE /dev/sd?1
SCSI device names can change if the hardware config changes, and you
want mdadm to find the devices no matter where they are.

NeilBrown