message from dmesg

[linux: ps_mon.cfg]

Hi All, i received these messagge by dmesg
i think they can be very dangerous
anyone can help me ?
what do they mean ?

thanks in advance
Fab

Major 09/15/03 02:17:54 pemihp03 UX10/dmesg OS LVM LVM: Path (device 0x1f039000) to PV 13 in VG 8 Failed!
Major 09/15/03 02:17:54 pemihp03 UX10/dmesg OS LVM LVM: Path (device 0x1f019300) to PV 16 in VG 8 Failed!
Major 09/15/03 02:17:54 pemihp03 UX10/dmesg OS LVM LVM: Path (device 0x1f039100) to PV 14 in VG 8 Failed!
Major 09/15/03 02:17:54 pemihp03 UX10/dmesg OS LVM LVM: Path (device 0x1f005400) to PV 0 in VG 8 Failed!
Major 09/15/03 02:17:54 pemihp03 UX10/dmesg OS LVM LVM: Path (device 0x1f005500) to PV 1 in VG 8 Failed!

[Mark Grant]

Looks like you have a hardware issue with a disk controller or SAN switch. I would expect that a vgdisplay -v of your volume groups might give a little more information

[Stefan Farrelly]

Youve lost connection so some of your disks. Contact your storage expert or supplier to investigate ASAP.

[Pradeep_3]

Hi,

Could you check the ioscan output
and look for all the devices with the
state as "NO_HW".

Is it a Direct attached storage or thru
SAN switch?

Regards
Fenil

[Steven E. Protter]

You are correct concerning danger.

If connected by fiber, you may get additional info from this command:

fcmsutil /dev/td0

replace td0 by the actual address of the card.

A physical inspection of the fiber card itself for lights, connections and such is warranted, because the hardware folks will ask you that anyway. Also a look at the lights on the fabric switch and the console for the disk array would be very helpful.

It is possible that the disk admin(hopefully not yourself) has changed zoning or ownership on the disks, or the fabric switch itself has failed. It depends how you are connected.

Depending on whats actually on those disks, I'd be wary of whether production is going to be normal today.

SEP

[Zeev Schultz]

5 LUNS in vg8. I'd look either for a controller
or a storage device fault (can be checked if you have more LUNs from this storage used ).
"ioscan -fn" to see state (NO_HW,CLAIMED,UNKNOWN).Seek "scsi" errors in syslog.log , if found could also lead to physicall / logical scsi errors (ie termination,cable,bad set scsi id).

Usually most storage arrays provide specific commands to check storage state.

Zeev