SUMMARY: error messages from hard disk

On Mon, Aug 28, 2006 at 11:20:40AM +0300, Athanasios Douitsis wrote:

Hi everyone,

I got this from one of my machines:

scsi: [ID 107833 kern.warning] WARNING: /pci@1f,4000/scsi@3/sd@8,0 (sd7):
Error for Command: read(10) Error Level: Retryable
scsi: [ID 107833 kern.notice] Requested Block: 18739904 Error Block: 44645507
scsi: [ID 107833 kern.notice] Vendor: SEAGATE Serial Number: 3HZ04FWP
scsi: [ID 107833 kern.notice] Sense Key: Media Error
scsi: [ID 107833 kern.notice] ASC: 0xc (write error), ASCQ: 0x0, FRU: 0xd3
scsi: [ID 107833 kern.warning] WARNING: /pci@1f,4000/scsi@3/sd@8,0 (sd7):
Error for Command: read(10) Error Level: Retryable
scsi: [ID 107833 kern.notice] Requested Block: 18739904 Error Block: 44645510
scsi: [ID 107833 kern.notice] Vendor: SEAGATE Serial Number: 3HZ04FWP
scsi: [ID 107833 kern.notice] Sense Key: Hardware Error
scsi: [ID 107833 kern.notice] ASC: 0x44 (internal target failure), ASCQ: 0x0, FRU: 0xd3


This disk is one of the two mirrors on a RAID-1 setup using SVM. Metacheck says everything ok.

Is this a preamble to a complete failure of the disk, or just a bad block? Should I ignore or
have the disk replaced?

Many thanks in advance for any responce,

Cheers,
Athanasios
_______________________________________________
sunmanagers mailing list
sunmanagers@xxxxxxxxxxxxxxx
http://www.sunmanagers.org/mailman/listinfo/sunmanagers

Many thanks to everyone for your speedy reply,

The general concensus is that this kind of errors probably signify that some block has gone bad inside
the disk. Some managers have suggested to use the "format" command with either the "defect" subcommand to count
the defects or the "analyze" subcommand to possibly make repairs. However to use "analyze" the disk must be
unmounted. Since this is a disk that is a submirror, detaching it and reattaching it to SVM may prove more work
than its worth it.

Most of the replies say that if the problem reappears several times, it will be a good time to change the disk. But
if this is a single occurence, it can be ignored, at least temporarily.

Kind regards,
Athanasios
_______________________________________________
sunmanagers mailing list
sunmanagers@xxxxxxxxxxxxxxx
http://www.sunmanagers.org/mailman/listinfo/sunmanagers

Relevant Pages

  • Re: 280R + SVM: Controller or disk problem ?
    ... > The system is under maintenance contract for 3 years. ... > replacement disk but I need the procedure to replace it I really don't ... > mirrors and recreate submirror after having replace the disk and copied the ... The SVM doc is not very useful at this point as they don't ...
    (comp.unix.solaris)
  • Re: SVM: 2 or 3 metadb/disk?
    ... Here's what I would like SVM to do in the standard 2-disk config which ... One system disk fails. ... SVM enters a degraded mode (setting some suitable flag in the ... Depending on HW machine either carries on, or reboots, in degraded ...
    (comp.sys.sun.admin)
  • Re: ZFS hot swapping and OS hangs
    ... A disk in A5200 or in the U60? ... the reliability of SVM. ... also recover if only a parity block is bad. ... disks in a RAID5 SVM volume and had 2 hot spares. ...
    (comp.unix.solaris)
  • Question about solaris volume manager
    ... I am trying to install solaris volume manager for the first time. ... SVM components are analogous to vxvm components? ... D1000 and the second internal disk on the 420. ...
    (comp.unix.solaris)
  • Re: 280R + SVM: Controller or disk problem ?
    ... > 1) first verify that all your failed slices are on the same physical disk. ... > SVM is reporting mirror failures on both sides of mirrors, ... Only one slice failed. ...
    (comp.unix.solaris)