topic Re: specific disk failed problem,,again in Operating System - HP-UX

specific disk failed problem,,again

KCS_1 — Tue, 31 Dec 2002 03:08:32 GMT

Re: specific disk failed problem,,again

KCS_1 — Tue, 31 Dec 2002 03:15:31 GMT

sorry,,i made a mistake of under parts..

I changed my server's internal disk(not bootable disk) cause by when i was running "ioscan" command received "NO_HW"

Re: specific disk failed problem,,again

T G Manikandan — Tue, 31 Dec 2002 04:31:48 GMT

Disk is not responding for the scsi request.
Thus,the scsi resets after searching for some time.

Definitely you need to look at the scsi connections and the //scsi termination//.

Revert for the results

Re: specific disk failed problem,,again

Glerick Dale — Tue, 31 Dec 2002 11:54:01 GMT

The problem your having is with your SCSI. Either the cable from the disk drive to the scsi or the scsi it's self. Change the cable first. If this does not work try connecting disk drive on another scsi if one is availble.

Re: specific disk failed problem,,again

Eugeny Brychkov — Tue, 31 Dec 2002 12:33:31 GMT

disk 1 8/4.9.0 sdisk CLAIMED DEVICE SEAGATE ST39173WC
/dev/dsk/c0t8d0 /dev/rdsk/c0t8d0
You receive error on this D380 internal disk.
I see that device file is wrong. Disk has SCSI Id=9, but looking to device file it should be 8. Valid device file for disk at path 8/4.9.0 should be /dev/[r]dsk/c0t9d0. Solution: do rmsf for this special file, and then issue 'insf -e'. But if you have this disk in vg then procedure should be more complex (up to vg killing).
Post /dev/dsk and /dev/rdsk directory listing; 'lssf /dev/rdsk/c0t8d0' output; 'pvdisplay -v /dev/c0t8d0' output (in case this disk is LVM disk)
Eugeny

Re: specific disk failed problem,,again

Anthony deRito — Tue, 31 Dec 2002 15:36:40 GMT

If the problem is accompanied by a message about pv[#] returned to vg[#], then the error can usually be attributed to a low timeout value on the disk driver. By default, the timeout is 30 seconds.

Increase the timeout up to the maximum of 180 seconds:
pvchange -t 180 /dev/dsk/disk_device
Increasing the timeout will not affect I/O performance on the disk.

Re: specific disk failed problem,,again

KCS_1 — Wed, 01 Jan 2003 02:37:07 GMT

Sorry~ Eugeny Brychkov

I attached wrong infomation about disk target so modified to new things

==OLD==

disk 1 8/4.9.0 sdisk CLAIMED DEVICE SEAGATE ST39173WC
/dev/dsk/c0t8d0 /dev/rdsk/c0t8d0

==New==

disk 1 8/4.9.0 sdisk CLAIMED DEVICE SEAGATE ST39173WC
/dev/dsk/c0t9d0 /dev/rdsk/c0t9d0

Happy new year

Re: specific disk failed problem,,again

Steven E. Protter — Wed, 01 Jan 2003 05:53:18 GMT

Understand one thing.

Your disk is dead. If it still lives, get the data off, it will die soon enough.

You will generate an lbolt on a D class box if you swap out a hot swapable drive. The only way to correct it is to restart the box.

Let me tell you about the possible problems that could contribute to this issue.

1) An unused, trully dead drive. If a drive is totally unused, no raw disk areas, no lvols, it can be dead and affect other drives in the drive cage.

2) Bad drive cage. He had one fail a few months back, it was a paint to diagnose. It kept destroying drives. A drive cage on a D box provides power and scsi connection to the hot swappable, internal drives.

3) Bad SCSI card or cable. In a D box with a drive cage you can have a bad cable and it kills drives.

For certain, you need hardware on site, even if you have to pay the hourly. Have them come with a replacement drive, drive cage and cables. Let them take the time to inspect the system thoroughly.

Do not settle for a drive replacement on a problem that keeps coming back. Its probably not just the drive.

I currently am responsible for two boxes, a D320 and a D380. AS far as drive problems go, if it eats two in a year, thats very unusual. These drives should outlast us both.

Been there, done that, go the tee shirt. In fact, a closet full.

Good luck sir.