LVM or disk problem?

[Bryan Barcelo]

Hi,

I got this error message from the server:
LVM: PV 0 has been returned to vg[2]
LVM: vg[2]: pvnum=0 (dev_t=0x....) is POWERFAILED
Disk at 56/52.4.0 is not responding. Check device, power, and cables.

I already checked and all the disks were powered-up and mounted. Anyone know what's happening?

TIA

[PIYUSH D. PATEL]

Hi,

This could be a temporary problem. The disk might not be getting sensed and hence this message. Are your disks connected to the disk array thro fiber channel cards. Sometimes for the old fiber channels cards, I used to get these messages. Do you have alternate path configured ??

You have to keep on checking the syslog.log files for this error to happen again. It is surely not an LVM problem. It is related to the hardware only....some loose connection etc.

Check all the devices and rectify loose conenctions.

Piyush

[Christopher McCray_1]

Hello,

did you try diskinfo?

# diskinfo /dev/rdsk/c?t?d?

It should return information immediately.

also you can use dd

# dd if=/dev/rdsk/c?t?d? of=/dev/null

Hope this helps

Chris

[MANOJ SRIVASTAVA]

Hi Bryan


This refers to the second disk in the volume group , this disk is bad , please do a diskinfo /dev/dsk/cxtydz and check for the status , I hhope you ahve a good back for the volume group.


Manoj Srivastava

[PIYUSH D. PATEL]

Hi,

If you are using AutoRaid 12H then these could be a timeout problem also.

Some of these errors can be reduced, masked and/or eliminated by changing the I/O timeouts from the default to 180 seconds or more. In general, however, the default value is quite adequate to handle normal drive seek errors, so increasing this value serves mainly to mask the fact that the disk drives have one or more bad spots that are not in the drive defect tables.

Do diskinfo and ioscan on the disk and check out for the errors.

Piyush

[Shahul]

Hi

Did U checked up for an HDD error in /var/adm/syslog/syslog.log file? Please check it up..

And please follow these

#strings /etc/lvmtab

Not down the device path of HDD's, Then

#ioscan -fnC disk

Make sure that all HDDs are listed here.. Then

#diskinfo /dev/rdsk/cxtxdx

Do this for all HDDs

By this time U will be able to trace the problematic HDD.

Best of luck
Shahul

[Sanjay_6]

Hi Bryan,

Looks like an intermediate disk problem. It could be a disk timeout problem. You can solve this by increasing the disk timeout, but i would suggest you to replace the disk. If this is an autoraid lun, you can do a logprint and then look at the array log and determine which disk is timing out and replace that disk.

To increase the timeout, try

pvchange -t 180 /dev/dsk/??????

The timeout can be upto 300 seconds. By defaults it is 30, i think.

I have always preffered replacing the disk rather than incresing the timeout.

[John Dvorchak]

I periodically get these messages on my older systems (G/H/I) that are connected to and EMC array. HP suggegest that I increase ALL disks in the VG to 180 timeout as was stated earlier. If you just increase the timeout for one disk in a VG the errors will continue.

pvchange -t 180 /dev/dsk/c_t_c_

Now this doesn't mean that you DON'T have a failing drive. So be carefull and perform all of the above suggested steps. I run sar also and am able to have the errors coinside with periods of heavy disk I/O so I feel confident that I am not losing drives.