LVM: vg[0]: pvnum=0 (dev_t=0x1f006000) is POWERFAILED ???

[cl_5]

Hi Forumers!

I've got some serious messages in my systemlog.
Could anybody please advise what to do?

Oct 28 02:00:23 pacmas02 vmunix: SCSI: Unexpected Disconnect -- lbolt: -593051251, dev: 1f006000, io_id: 90a02d
Oct 28 02:00:23 pacmas02 vmunix: SCSI: Target detected parity error -- lbolt: -593051151, dev: 1f006000
Oct 28 02:00:23 pacmas02 vmunix: SCSI: Target detected parity error -- lbolt: -593051019, dev: 1f006000
Oct 28 02:00:23 pacmas02 vmunix: SCSI: Target detected parity error -- lbolt: -593050953, dev: 1f006000
Oct 28 02:00:23 pacmas02 vmunix: SCSI: Target detected parity error -- lbolt: -593050951, dev: 1f006000
Oct 28 02:00:23 pacmas02 vmunix: SCSI: Target detected parity error -- lbolt: -593051151, dev: 1f006000
Oct 28 02:00:37 pacmas02 vmunix: SCSI: Request Timeout; Abort Tag -- lbolt: -593048191, dev: 1f006000, io_id: 90a02c
Oct 28 02:00:43 pacmas02 vmunix: SCSI: Request Timeout; Abort Tag -- lbolt: -593048091, dev: 1f006000, io_id: 90a030
Oct 28 02:00:43 pacmas02 vmunix:
Oct 28 02:00:43 pacmas02 vmunix: SCSI: Write error -- dev: b 31 0x006000, errno: 126, resid: 8192,
Oct 28 02:00:43 pacmas02 vmunix: ^Iblkno: 2487296, sectno: 4974592, offset: -1747976192, bcount: 8192.
Oct 28 02:00:43 pacmas02 vmunix:
Oct 28 02:00:43 pacmas02 vmunix: SCSI: Read error -- dev: b 31 0x006000, errno: 126, resid: 2048,
Oct 28 02:00:43 pacmas02 vmunix: ^Iblkno: 8, sectno: 16, offset: 8192, bcount: 2048.
Oct 28 02:00:43 pacmas02 vmunix: LVM: vg[0]: pvnum=0 (dev_t=0x1f006000) is POWERFAILED
Oct 28 02:00:43 pacmas02 vmunix: LVM: Recovered Path (device 0x1f006000) to PV 0 in VG 0.
Oct 28 02:00:43 pacmas02 vmunix: LVM: Restored PV 0 to VG 0.
Oct 28 02:16:11 pacmas02 vmunix:
Oct 28 02:16:11 pacmas02 vmunix: SCSI: Read error -- dev: b 31 0x006000, errno: 126, resid: 2048,
Oct 28 02:16:11 pacmas02 vmunix: ^Iblkno: 8, sectno: 16, offset: 8192, bcount: 2048.
Oct 28 02:16:11 pacmas02 vmunix: LVM: vg[0]: pvnum=0 (dev_t=0x1f006000) is POWERFAILED
Oct 28 02:16:16 pacmas02 vmunix: LVM: Recovered Path (device 0x1f006000) to PV 0 in VG 0.
Oct 28 02:16:16 pacmas02 vmunix: LVM: Restored PV 0 to VG 0.
Oct 28 02:24:23 cas020 xntpd[11133]: synchronisation lost
Oct 28 02:24:36 cas107 xntpd[1376]: synchronisation lost
Oct 28 02:26:04 pacmas02 vmunix: LVM: Recovered Path (device 0x1f006000) to PV 0 in VG 0.
Oct 28 02:26:09 pacmas02 vmunix: LVM: Restored PV 0 to VG 0.
Oct 28 02:26:09 pacmas02 vmunix:
Oct 28 02:26:09 pacmas02 vmunix: SCSI: Async write error -- dev: b 31 0x006000, errno: 126, resid: 8192,
Oct 28 02:26:09 pacmas02 vmunix: ^Iblkno: 1391648, sectno: 2783296, offset: 1425047552, bcount: 8192.
Oct 28 02:26:09 pacmas02 vmunix: LVM: Recovered Path (device 0x1f006000) to PV 0 in VG 0.
Oct 28 02:26:09 pacmas02 vmunix: LVM: Restored PV 0 to VG 0.
Oct 28 02:26:25 pacmas02 vmunix:

Thanks a lot!
cl

[Sridhar Bhaskarla · Accepted solution]

Hi,

Looks like your c0t6d0 has been having intermittent issues. Do a 'vgdisplay -v' and see if you see any errors. Run 'ioscan -f' and make sure you don't have NO_HW corresponding to the hardware path of c0t6d0. It looks like your root disk with a mirror. If you keep seeing these messages, you will need to have it replaced. To decode the device, use this formula

0x1f006000

1f=major number for sdisk driver.
00=instance number for this controller
6=target
0=LUN
00=ignore for disk device

Or simply do a 'll /dev/dsk |grep 006000'

Another test is to run a dd on it while watching your syslog.log

tail -f syslog.log
dd if=/dev/rdsk/c0t6d0 of=/dev/null bs=1024k (in another window)

-Sri

[john korterman]

Hi,
your disk c0t6d0 may be dead. Check the disk by this command:
# ioscan -fnC disk
and look for the status for the disk in question. If it say NO_HW - and you know for sure that noone has switched off the power for taht particular disk - call HP and have it replaced.
The disk belongs to vg00. Check which logical volume by this command:
# vgdisplay -v /dev/vg00|more
and see where something is stale. I assume that you use mirroring, and the system is still running omn the remaining disk. But I think you should report the error to HP.

regards,
John K.

[cl_5]

Thanks everybody for help!

There are no mirrors. This is just a single rootdisk.

I've done a vgdisplay -v vg00. No errors
lvdisplay -v on all lvols. no stale blocks
ioscan -fnCdisk. Device claimed.
After reboot no error messages appeared. Everythings seems to alright.

Maybe a harbinger to a hdd crash?