Logical Volume Problem

[hahmedma]

Hello all,

   i had a problem on creating the mirror for te following 2 logical volume (lvol7 and lvol13) thier data as follow:

   LV Name                     /dev/vg00/lvol7                     
   LV Status                   available/syncd
   LV Size (Mbytes)            3776
   Current LE                       59
   Allocated PE                   59
   Used PV                          1  

   LV Name                     /dev/vg00/lvol13
   LV Status                   available/syncd
   LV Size (Mbytes)            44480
   Current LE                  695
   Allocated PE                695
   Used PV                     1

 --- Physical volumes ---
   PV Name                     /dev/dsk/c0t6d0s2
   PV Status                   available
   Total PE                    2173
   Free PE                     34
   Autoswitch                  On

   PV Name                     /dev/dsk/c6t6d0s2
   PV Status                   available
   Total PE                    2173
   Free PE                     764
   Autoswitch                  On


   --- Physical volume groups ---
   PVG Name                    pvg0
   PV Name                     /dev/dsk/c0t6d0s2

   PVG Name                    pvg1
   PV Name                     /dev/dsk/c6t6d0s2

and when i check the PV for each hard there is NO Stable existed as follow:

 pvdisplay -v /dev/dsk/c6t6d0s2 | grep stale------> no stale existed

 pvdisplay -v /dev/dsk/c0t6d0s2 | grep stale------> no stale existed

when i try to create the mirror for above lvols as follow it result is fails as follow:

lvextend -m 1 /dev/vg00/lvol7 /dev/dsk/c6t6d0s2
The newly allocated mirrors are now being synchronized. This operation will
take some time. Please wait ....
lvextend: Couldn't re-synchronize stale partitions of the logical volume:
I/O error
lvextend: Couldn't resynchronize logical volume "/dev/vg00/lvol7".

and i had check the system log i had find the following EMS releated to disk as follow:

Jun 24 13:07:50 mbnlpdh2 EMS [4465]: ------ EMS Event Notification ------   Value: "CRITICAL (5)" for Resource: "/storage/events/disks/default/0_0_0_3_0.6.0"     (Threshold:  >= " 3")    Execute the following command to obtain event details:   /opt/resmon/bin/resdata -R 292618242 -r /storage/events/disks/default/0_0_0_3_0.6.0 -n 29261
8242 -a

when i run it the result as follow:

CURRENT MONITOR DATA:

Event Time..........: Sun Jun 24 14:14:19 2012
Severity............: CRITICAL
Monitor.............: disk_em
Event #.............: 101037
System..............: mbnlpdh2

Summary:
     Disk at hardware path 0/0/0/3/0.6.0 : Media failure


Description of Error:

     The device was unsuccessful in reading or writing data for the current I/O
     request due to a data synchronization mark error on the medium.

Probable Cause / Recommended Action:

     Reformatting the medium may fix the problem.

     Alternatively, the medium in the device is flawed. If the medium is
     removable, replace the medium with a fresh one.

     Alternatively, if the medium is not removable, the device has experienced
     a hardware failure. Contact your HP support representative to have the
     device checked.

so i hope i had explain everything for this problem to to create the Mirror for logical volume...does there is any failure on the hard desks  need to replace.

also i run the following 2 command as follow:

dd if=/dev/rdsk/c0t6d0s2 of=/dev/null bs=1024k
3308+0 records in
3308+0 records out
 dd if=/dev/rdsk/c6t6d0s2 of=/dev/null bs=1024k
15578+0 records in
15578+0 records out

i hope to be clear everything to you....waiting your feedack..

[Bill Hassell]

Your dd test:

dd if=/dev/rdsk/c0t6d0s2 of=/dev/null bs=1024k
3308+0 records in
3308+0 records out
 dd if=/dev/rdsk/c6t6d0s2 of=/dev/null bs=1024k
15578+0 records in
15578+0 records out

shows that c0t6d0 is way too small. The I/O error indicates that you have a disk failure and the EMS message confirms that condition.

>> no stale existed

The status "stale" does not mean anything about the health of the disk. In your case, the problem areas have not been mirrored yet so they can't be stale, they are simply unassigned. Your disk needs to be replaced and the entire mirroring process started again.

[Matti_Kurkela · Accepted solution]

...and after the replacement of the c0t6d0 disk, the data of the non-mirrored LVs lvol7 and lvol13 must be restored from backups.

If you don't have up-to-date backups, you should attempt to backup the contents of lvol7 and lvol13 as soon as possible, with the understanding that some of the data is very likely already lost. As you have not noticed the data loss before this situation, the data that is lost might be infrequently used, and so you might be able to reconstruct the data by recovering the missing files from an older backup... so don't overwrite or discard any old backups of lvol7 or lvol13.

[hahmedma]

thanks for your support...now does i can take a full backup for the system  as offline or theris a problem...and restore it on the new media after replaced...then create mirror after that...

[Matti_Kurkela]

Your dd test already indicated that some parts of the c0t6d0 disk are unreadable, and the EMS notification confirmed that. So yes, you have a problem.

If you are very lucky, all the files currently on lvol7 and lvol13 are not located in the unreadable spots and you can complete the backup. But if some data is located in an unreadable spot on the disk, the backup program won't be able to read it: it will either give up with an error message, or produce a backup containing damaged data.

If you don't already have a backup of your data, it is possible that some of your data is already destroyed and cannot be recovered by taking a new backup now. The only way to know how much damage is already done is to copy whatever you can to a safe location and then verify as much of the data as you can.

If the data is critical to you, you might want to stop using the disk right now and contact data recovery professionals (companies like Kroll Ontrack, for example). It might be expensive, but maybe less expensive than losing your critical data.