Stripset on HSG80

[Ian Derringer]

Greetings,
I have a stripset consist of 3 x 18.0 GB drives. I had a mountverify on this volume and now the volume is off-line. I am not able to mount the volume. Therefore, I connected through VMS "set host/scsi $1$dga21", I can see the stripeset and the drives that associate with the stripeset. However, one (1) displayed "No Information Available" and "Not Yet Known". I tried to reset the errors by "clear_errors disk4100 UNKNOWN" but no luck. Is the a way for me to bring this drive back on-line without losing the data on this array?? Any ideas how I can bring this volume to live and mount it, so that I can do a quick backup before replacing the bad drive?? Please help!!

Thank you in advance.

Ian D.

[Uwe Zessin · Accepted solution]

On the HSG, a disk drive entry does not disappear by magic if the disk drive breaks. I would try to 'restart' the disk drive:

- pull it a bit from the bay
- wait a minute
- put it back in
- see if it works now
If that does not work, I am afraid you have to exercise your desaster recovery plan.

[Ivan Ferreira]

you should post the results of

show disks
show units
show storage
show stripe
show failed
show DXXX, where DXXX is the unit for the stripeset
show stripe SXXXX, where SXXX is the stripe set name.

And you cleared the error of one disk, not from the stripe set.

[Ian Derringer]

Hi,
Here are the results;
DISK13S1 stripeset DISK41300 D13
DISK51300
DISK61200
Switches:
CHUNKSIZE = 64 blocks
State:
INOPERATIVE
DISK51300 (member 1) is NORMAL
DISK61200 (member 2) is NORMAL
DISK41300, no information available
Size: NOT YET KNOWN


Name Storageset Uses Used by
------------------------------------------------------------------------------

FAILEDSET failedset
Switches:
NOAUTOSPARE


Name Storageset Uses Used by
------------------------------------------------------------------------------

DISK13S1 stripeset DISK41300 D13
DISK51300
DISK61200
Switches:
CHUNKSIZE = 64 blocks
State:
INOPERATIVE
DISK51300 (member 1) is NORMAL
DISK61200 (member 2) is NORMAL
DISK41300, no information available
Size: NOT YET KNOWN



Let me know.

Thanks so much!!
Ian

[Ian Derringer]

One question; How do you clear the error on one (1) drive?

Ian

[Uwe Zessin]

You can't, sorry.

If the controller can no longer talk to the disk because it's simply dead, then you have to delete the unit, delete the stripeset, replace the physical disk drive. Now you need to recreate the stripeset and run a new INITIALIZE so that all disks receive a metadata update. Next, you have to set up the unit and SSP. Final task is to restore from backup.

[generic_1]

It sounds like you were not using raid protection. You may want to conisder this if you are not.

[Uwe Zessin]

That's correct. A 'HSG Stripeset' implements RAID-0.

[Ivan Ferreira]

Your stripset is lost, RAID-0 does not have fault tolerance.

DISK13S1 stripeset DISK41300 D13
DISK51300
DISK61200
Switches:
CHUNKSIZE = 64 blocks
State:
INOPERATIVE
DISK51300 (member 1) is NORMAL
DISK61200 (member 2) is NORMAL
DISK41300, no information available
Size: NOT YET KNOWN


"DISK41300, no information available"

The disk41300 is missing, you may need to replace it, and the stripeset is deactivated because a missing disk.

You should consider using a RAID-5 instead.

On the HSG, use the

run config command

and then post the results of:

show disk41300
show DISK13S1
show D13

[Ian Derringer]

I think you guys correct about the drive being dead or?? Anyways, I will replace these drives with 2 x 72GB drives RAID0 + 1

What would you guys think of my decision??

Thanks for all your help!!

Ian