12h autoraid - missing disk & recovery failed warnings

[Ben Hohman]

We had a 12H crash and upon restarting it would hang. We replaced the disk in A3 and the array added the drive and rebuilt (auto-include, auto-rebuild auto-hotspare are all enabled). Since then we have two errors: missing disk, recover failed reported by arraydsp -s. If we try to rebuild the array (arrayrbld -r), we are told that no recovery is needed. Using mstm, I have excersised the array and it reports back ok. I think that the missing disk and recovery warnings are bogus. Is there any way to clear the warnings short of wiping the array?

[Steven E. Protter]

Shalom,

New disks need to be initialized with pvcreate

They also may need to be forced out of the volume group configuration prior to replacement.

http://docs.hp.com/en/5991-1236/When_Good_Disks_Go_Bad_WP.pdf

Found it outside the HP net as well.
http://www.deltago.co.kr/Board/Down.aspx?FileName=When%20Good%20Disks%20Go%20Bad.pdf

copyright problems?

SEP

[Ben Hohman]

Steven,

The disk in question is under the control of the array, it's not in the chassis of the server, so I don't see how the logical volume manager comes into play. No changes were made to the size of the volumes on the array or the server's disks.

Ben

[Denver Osborn]

Not speaking from experience on this, but after you replaced the disk in A3 and received the missing disk warnings, have you attempted to restart the 12H? (arraymgr shut/start)

-denver

[OldSchool]

see this link for the System Admin Guide for the 12H. check out pages around 52-53 for command line diagnostics, non-destructive tests of a disk in a specific slot and how to read / clear the hardware log in the array.

I suspect the monitor is picking up the death of the old drive from that log, but I can't confirm that (no more autoraids on site)

[OldSchool]

I guess the link would help????
http://h10032.www1.hp.com/ctg/Manual/lpg28365.pdf

[Ben Hohman]

Dennis,

We have shutdown the array using the array front panel.

Ben

[Ben Hohman]

OldSchool,

I had cleared the logs for the disk in slot A3 (the only one we replaced) as well as the controller.

ghpts1:/# arraylog -e PTSMAIN

Primary Controller = X

Controller X Vendor ID = HP
Controller X Product ID = C5447A
Controller X Firmware = F209030416
Controller X Serial Number = R85FADF979691111
Controller Y Vendor ID = HP
Controller Y Product ID = C5447A
Controller Y Firmware = F209030416
Controller Y Serial Number = R84TADF940971111
No Events in Log

ghpts1:/# arraylog -u PTSMAIN

Primary Controller = X

Controller X Vendor ID = HP
Controller X Product ID = C5447A
Controller X Firmware = F209030416
Controller X Serial Number = R85FADF979691111
Controller Y Vendor ID = HP
Controller Y Product ID = C5447A
Controller Y Firmware = F209030416
Controller Y Serial Number = R84TADF940971111

Usage Log
Total ECC Errors = 0
ghpts1:/#
***
I still get the warnings.

[OldSchool]

did you try "drivetest" (pg 52 in the linked doc?)

[Torsten.]

No idea if it really is, but it could be a firmware issue, because you are running hp54 and latest for this oldie was hp62.
Caution, if you ever find this firmware and consider an update, update the commands patch first!

[Ben Hohman]

OldSchool,

Yes, I did a drivetest and it passed.

Ben