Message syslog.log

[Jose_129]

I have a message syslog.log

Jul 12 05:35:43 pgcaxis above message repeats 5 times
Jul 12 05:35:43 pgcaxis vmunix: LVM: Recovered Path (device 0x1f094000) to PV 1 in VG 4.
Jul 12 05:35:43 pgcaxis vmunix: LVM: vg[1]: pvnum=1 (dev_t=0x1f083200) is POWERFAILED
Jul 12 05:35:37 pgcaxis vmunix: SCSI: Async write error -- dev: b 31 0x095400, errno: 126, resid: 4096,
Jul 12 05:35:43 pgcaxis vmunix: LVM: Restored PV 1 to VG 7.
Jul 12 05:35:43 pgcaxis vmunix: LVM: Recovered Path (device 0x1f085400) to PV 1 in VG 10.
Jul 12 05:35:44 pgcaxis vmunix: LVM: Restored PV 1 to VG 4.
Jul 12 05:35:45 pgcaxis vmunix: LVM: vg[7]: pvnum=1 (dev_t=0x1f084600) is POWERFAILED
Jul 12 05:35:45 pgcaxis vmunix: LVM: vg[10]: pvnum=1 (dev_t=0x1f085400) is POWERFAILED
Jul 12 05:35:43 pgcaxis vmunix: LVM: vg[1]: pvnum=1 (dev_t=0x1f083200) is POWERFAILED
Jul 12 05:35:45 pgcaxis vmunix: LVM: Recovered Path (device 0x1f084400) to PV 1 in VG 6.
Jul 12 05:35:45 pgcaxis vmunix: LVM: vg[6]: pvnum=1 (dev_t=0x1f084400) is POWERFAILED
Jul 12 05:35:45 pgcaxis vmunix: LVM: Restored PV 1 to VG 10.
Jul 12 05:35:46 pgcaxis vmunix: LVM: Restored PV 1 to VG 6.
Jul 12 05:35:50 pgcaxis vmunix: LVM: Recovered Path (device 0x1f093200) to PV 1 in VG 1.
Jul 12 05:35:50 pgcaxis vmunix: LVM: Performed a switch for Lun ID = 0 (pv = 0x000000004c01f040), from raw device 0x1f083200 (with priority: 1, and current flags: 0xc0) to raw device 0x1f093200 (with priority: 0, and current flags: 0x80).
Jul 12 05:35:50 pgcaxis vmunix: LVM: Recovered Path (device 0x1f083200) to PV 1 in VG 1.
Jul 12 05:35:54 pgcaxis vmunix: LVM: Restored PV 1 to VG 1.
Jul 12 05:37:45 pgcaxis vmunix: LVM: Recovered Path (device 0x1f094400) to PV 1 in VG 6.
Jul 12 05:37:45 pgcaxis vmunix: LVM: Performed a switch for Lun ID = 0 (pv = 0x000000004c30d840), from raw device 0x1f084400 (with priority: 1, and current flags: 0x0) to raw device 0x1f094400 (with priority: 0, and current flags: 0x80).
Jul 12 05:39:37 pgcaxis vmunix: LVM: Recovered Path (device 0x1f094600) to PV 1 in VG 7.
Jul 12 05:39:37 pgcaxis vmunix: LVM: Performed a switch for Lun ID = 0 (pv = 0x000000004c34e040), from raw device

What problem is?

regards,

Jose

[RAC_1]

You have problem with disk - /dev/dsk/c15t9d4
(lies in vg07)

The message is telling, that OS had problem talking to disk and again recovered.

Run diskinfo/stm on the disk and check.
Also, if you alternate paths for this disk, add it.

Anil

[Jose_129]

and the message:

Jul 12 05:35:43 pgcaxis vmunix: LVM: Recovered Path (device 0x1f094000) to PV 1 in VG 4.
Jul 12 05:35:43 pgcaxis vmunix: LVM: vg[1]: pvnum=1 (dev_t=0x1f083200) is POWERFAILED

[RAC_1]

0x1f083200 -- disk failed. /dev/dsk/c31t8d2

(grep "0x1f083" /dev/dsk/*) -- This is the disk that has problems.
(powerfailed-stopped spinning)
Any visual indication, on the disk, that has gone bad??

Anil

[Muthukumar_5]

hai,

Problem because of power failure or network disconnection (physical wires disconnection).

Same error and discussion here at,
http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=620061

Regards,
Muthukumar.

[Jeff Schussele]

Hi Jose,

You're getting indications of either:

1) Time outs down your c8 & c9 paths. I'd suspect that you need to increase the time out value from 30 seconds up to 60 or 90
pvchange -t or lvchange -t

2) Intermittent problems with eithet the c8 or c9 controller.

I stongly suspect the former & not the latter.

[Jeff Schussele]

Hi RAC,

When the hex "x" is in front of "1f" then it's a device number - in this case the major number for disk devices.
Then the next 2 are controller - in hex. The next is target - in hex & the final is the LUN in hex or decimal as it's never > 7.

So above is c8t5d4 c9t5d4 c8t4d6 c8t3d2 c8t4d4 etc.....

Rgds,
Jeff

[Mohanasundaram_1 · Accepted solution]

Hi Jose,

Jeff is correct about increasing the pvtimeout value and about the device file too. the 1F hex number indicates the Major number of the disk devices (31).

I presume you are having a VA7xxx Storage. It is worthwhile to verify if the controllers, swithces or cables are not having any issues.

If your VA firmware is pretty old, try updating it to the latest. If the storage is indeed VA, logprn should give more details about the happenings on this array.

My Action plan would be,

1) pvchange -t 90 /dev/
2) observe the syslog.log for errors.
3) Perform a visual check on the array for any amber LED.
4) Check the logprn for the history of the problem.
5) armdsp -a can give you the general status of the array.

Provide more details about the storage and topology for more hints on this problem.

Cheers,
Mohan.