Disk failure?

[marko asplund]

hi

I've a rp3410 system running HP-UX 11i. All the logical volumes are mirrored (using HP LVM + MirrorDisk/UX i believe) and there's a hotspare as well.

Now when i view the volume group info through SAM i see the following output:

Hw path disk Mb usable mb physical vg disk type status
0/1/1/0.0.0 70007 70007 -- Normal Failed, Spared to 0/1/1/1.2.0
0/1/1/0.1.0 70007 70007 -- Normal Available
0/1/1/1.2.0 70007 70007 -- Spare Active, Sparing 0/1/1/0.0.0

How should this be interpreted? Is it a sign of disk failure?

If it's a disk failure, how do i replace the disk so that it will be made part of the mirrored volume?


aspa

[Chan 007]

Hi,

Try ioscan -fCdisk,

Also check dmesg and /var/adm/syslog/syslog.log for failures errors.

Log a call with HP, they will sort that out.

Yes you lost one disk.

007

[James R. Ferguson · Accepted solution]

Hi:

Yes, it appears that you have had a disk failure, but that you have excellent high-availability protection because of "sparing" -- a hot-standby spare disk that is "taken-over" to provide redundant data protection.

The "Managing Systems and Workgroups" guide provides more information on this feature:

http://docs.hp.com/en/B2355-90912/B2355-90912.pdf

As for replacement procedures, the newest, most comprehensive guide is this whitepaper:

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

Regards!

...JRF...

[marko asplund]

hi

thanks for your quick replies 007 and JRF.

here's what ioscan says:

nebbiolo# ioscan -fCdisk
Class I H/W Path Driver S/W State H/W Type Description
=========================================================================
disk 0 0/0/2/0.0.0.0 sdisk CLAIMED DEVICE TEAC DV-28E-C
disk 1 0/1/1/0.0.0 sdisk CLAIMED DEVICE HP 73.4GMAU3073NC
disk 2 0/1/1/0.1.0 sdisk CLAIMED DEVICE HP 73.4GMAU3073NC
disk 3 0/1/1/1.2.0 sdisk CLAIMED DEVICE HP 73.4GMAU3073NC


aspa

[Steven E. Protter]

Shalom Marco,

Assuming this system only had three disks to start, here is what probably happened.

disk 0 either failed or become unresponsive long enough for disk 2.0 the hot spare to take over.

Now it seems from ioscan that disk 0 is back. I'd tend not to believe without good hardware diagnostics, say a visit from HP hardware that the disk is going to continue working very long.

You can get your own look at the situation with mstm cstm or xstm, your choice, but I would not trust disk 0.

SEP

[Chan 007]

Mark,

Based on your IOSCAN, I think your disk is back. Better your syslog for failure. Also if you have ems configured you can see output in syslog during that time. Run that ems command see.

Better log a call with HP

007

[lawrenzo]

as mentioned before will be worth raising a call however you can test the disk as below:

run ioscan and get the disk id of disk 0

#ioscan -fnC disk

# diskinfo /dev/rdsk/c2t6d0

this should display something like:

SCSI describe of /dev/rdsk/c2t6d0:
vendor: HP 36.4G
product id: ST336706LC
type: direct access
size: 35566480 Kbytes
bytes per sector: 512

also use dd - an error will be diplayed if there is a problem with the disk, if not the command prompt appears:

# dd if=/dev/rdsk/ of=/dev/null bs=64k


HTH

[Sยภเl Kย๓คг]

Hi,

Let us first understand, how many disks are there in the system.

With the ioscan o/p, we see all detected disks are available to the system, assuming that you have not taken a reboot of the system after the failure.

Give a 'diskinfo' to all disks, use 'dd' to check for any possible media errors on all disks.

It is highly recommended to log call with HP and let the HP Engr to have look at physical connections and health check of the system.

Regards,
Sunil

[marko asplund]

hi

Thanks for all your helpful replies.

I'll contact HP to get someone diagnose the disk or a replacement.

The system has 3 disks installed and all of them are visible with ioscan. ioscan & diskinfo don't appear to have anything out of the ordinary to say:

nebbiolo# ioscan -fnC disk
Class I H/W Path Driver S/W State H/W Type Description
=========================================================================
disk 0 0/0/2/0.0.0.0 sdisk CLAIMED DEVICE TEAC DV-28E-C
/dev/dsk/c0t0d0 /dev/rdsk/c0t0d0
disk 1 0/1/1/0.0.0 sdisk CLAIMED DEVICE HP 73.4GMAU3073NC
/dev/dsk/c2t0d0 /dev/rdsk/c2t0d0
disk 2 0/1/1/0.1.0 sdisk CLAIMED DEVICE HP 73.4GMAU3073NC
/dev/dsk/c2t1d0 /dev/rdsk/c2t1d0
disk 3 0/1/1/1.2.0 sdisk CLAIMED DEVICE HP 73.4GMAU3073NC
/dev/dsk/c3t2d0 /dev/rdsk/c3t2d0

nebbiolo# diskinfo /dev/rdsk/c2t0d0
SCSI describe of /dev/rdsk/c2t0d0:
vendor: HP 73.4G
product id: MAU3073NC
type: direct access
size: 71687369 Kbytes
bytes per sector: 512

nebbiolo# diskinfo /dev/rdsk/c2t1d0
SCSI describe of /dev/rdsk/c2t1d0:
vendor: HP 73.4G
product id: MAU3073NC
type: direct access
size: 71687369 Kbytes
bytes per sector: 512

nebbiolo# diskinfo /dev/rdsk/c3t2d0
SCSI describe of /dev/rdsk/c3t2d0:
vendor: HP 73.4G
product id: MAU3073NC
type: direct access
size: 71687369 Kbytes
bytes per sector: 512

I'll post dd output when the commands finish (it seems to take a while).


aspa

[marko asplund]

The dd command finished ok for disks 1 and 2 but not for disk 0. The process running the command froze and it doesn't get terminated with sigkill either.

A few minutes after i started the dd for disk 0 i received a disk failure alert message from EMS for disk 0.


aspa