I tried Google.  Most of the responses - not surprisingly - reference
older packages such as RAID-Tools.  After scanning dozens of articles,
the only one I found which really addresses the issue uses the same
method I tried.  I know from experience this method works, but in
order for it to work one must be able to write to /sys/block/md0/md/
dev-sd*/slot, and the system won't let me do that.  I'm assuming this
is because the array is not running - catch 22.  I tried "mdadm --
incremental --rebuild --run /dev/md0", but I get a segmentation fault.

On Dec 23, 2:31 am, "Chris Miller" <[email protected]>
wrote:
> I suggest using Google.  I would imagine that with so many IBM
> Deathstar drives in this world that someone would have run into your
> problem before.
>
> On Mon, Dec 22, 2008 at 11:48 PM, [email protected]
>
>
>
> <[email protected]> wrote:
>
> > In that case, your problem exceeds the level of my experience, My
> > apology.
>
> > Do please let us know your solution when you find it.
>
> > On Dec 22, 11:39 pm, lrhorer <[email protected]> wrote:
> >> No, the drives are SATA, and it's a Silicon Image controller with port
> >> multipliers.  As I already mentioned, the array is using mdadm for
> >> management.  The controller is dumb as a post - no pun intended.
>
> >> On Dec 22, 10:21 pm, "[email protected]" <[email protected]>
> >> wrote:
>
> >> > Is this a SCSI array? If so what type of controller are you using?
>
> >> > My Compaq servers use a seperate "card specific" utility to set up the
> >> > array before the OS ever gets a look at it. I wonder if the DRAM on
> >> > your controller card has lost its programming during the crash.
>
> >> > On Dec 21, 2:28 pm, lrhorer <[email protected]> wrote:
>
> >> > > I am in serious need of some help.  I have a RAID 6 array running
> >> > > under Debian "Lenny" Linux using mdadm.  I have a 10 disc array set
> >> > > up, and I suffered a system failure requiring a reboot.  Now, the
> >> > > array won't start.  When I attempt to run --assemble, it thinks 3 of
> >> > > the drives are spare, so it won't run the array.  I tried manually
> >> > > assembling the array by writing to /sys/block/md0/md/dev-sd*/slot, but
> >> > > the system won't let me write to any of the directories, I assume
> >> > > because the array is not running.  How can I force mdadm to assemble
> >> > > and run at least 8 of the drives?  The issue which caused the drives
> >> > > to fail and then the system to halt has been resolved, but I can't
> >> > > figure out how to get the array back online without losing everything
> >> > > on the drives.  To answer the obvious question, I do have a backup
> >> > > system, but it failed 3 days ago, and it's going to be a while before
> >> > > the backup system can be repaired.
>
> >> > > If it's relevant, the drives are all unpartitioned, and the array is
> >> > > also unpartitioned, formatted as one large ext2 file system.
>
> --
> Registered Linux Addict #431495http://profile.xfire.com/mrstalinman| John 
> 3:16!http://www.fsdev.net/|http://lordsauron.wordpress.com/
> Parents, Take Responsibility For Your Kids!http://www.whattheyplay.com/
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Linux Users Group.
To post a message, send email to [email protected]
To unsubscribe, send email to [email protected]
For more options, visit our group at 
http://groups.google.com/group/linuxusersgroup
-~----------~----~----~----~------~----~------~--~---

Reply via email to