B2000 - SCSI Error - machine is slow.

[Konrad Hegner]

Hello

We work among other things with B2000 Workstation. One machine is now VERY slow and if I make the 'dmesg'-command the following message appears:


Nov 26 07:45
...
192,
blkno: 6757032, sectno: 13514064, offset: -1670733824, bcount: 8192.

SCSI: Async write error -- dev: b 31 0x025000, errno: 126, resid: 8192,
blkno: 6811344, sectno: 13622688, offset: -1615118336, bcount: 8192.

SCSI: Async write error -- dev: b 31 0x025000, errno: 126, resid: 2048,
blkno: 6791704, sectno: 13583408, offset: -1635229696, bcount: 2048.

SCSI: Async write error -- dev: b 31 0x025000, errno: 126, resid: 8192,
blkno: 6763144, sectno: 13526288, offset: -1664475136, bcount: 8192.

...

SCSI: Async write error -- dev: b 31 0x025000, errno: 126, resid: 8192,
blkno: 2412864, sectno: 4825728, offset: -1824194560, bcount: 8192.

SCSI: Read error -- dev: b 31 0x025000, errno: 126, resid: 2048,
blkno: 8, sectno: 16, offset: 8192, bcount: 2048.
LVM: vg[0]: pvnum=1 (dev_t=0x1f025000) is POWERFAILED

SCSI: Async write error -- dev: b 31 0x025000, errno: 126, resid: 16384,
blkno: 9625312, sectno: 19250624, offset: 1266384896, bcount: 16384.

SCSI: Async write error -- dev: b 31 0x025000, errno: 126, resid: 16384,
blkno: 9625328, sectno: 19250656, offset: 1266401280, bcount: 16384.

SCSI: Write error -- dev: b 31 0x025000, errno: 126, resid: 10240,
blkno: 2438, sectno: 4876, offset: 2496512, bcount: 10240.
LVM: Recovered Path (device 0x1f025000) to PV 1 in VG 0.
LVM: Restored PV 1 to VG 0.
LVM: Failed to automatically resync PV 1f025000 error: 5


And by the lvdisplay are 2 'stales' constant available. Also after a reboot.
The disk are connectet internal by the IDE-bus, and it is with mirroring.

Know anyone whats happen?

[Adisuria Wangsadinata_1 · Accepted solution]

Hi there,

From the message it looks your harddisk is get a problem (the boot disk - vg00 - c2t5d0). Better you call HP Response Center to log this problem, maybe the solution is replace the faulty disk.

Check also the mirror disk for stale state by running this command :

# lvdisplay -v /dev/vg00/

Please check whether the disk device with 'CLAIMED' status :

# ioscan -nfCdisk

You can check also the disk with 'dd' command, by trying to simulate the read-write on the particular disk :

# dd if=/dev/rdsk/cxtydz of=/dev/null bs=1024k

Hope this can help you.

Best Regards,
AW

[Steven E. Protter]

Most likely a disk is going. Best to get this thing backed up asap and prepare for total disk failure.

This kind of thing can be caused by bad cables, but that more likely would totally bring the system down.

vgdisplay -v /dev/vg00

If you have state areas mirroring has failed because of this issue.

SEP