recover VG with multiple disks

[George Liu_4]

A vg /dev/vg00 contains three disks, say /dev/dsk/c0t0d0, /dev/dsk/c0t1d0, and /dev/dsk/c0t2d0. Now, looks c0t0d0 is bad and not accessible. the other two are still good. we can still use vgimport to import them. There is any way to recover the contents from the good ones?

[RAC_1 · Accepted solution]

Does pvdisplay returns any thing?? (for bad disk)

Can you activate the VG?? (vgchange -a y -q n vgxx)
Do you what partitions (lvs) were residing on the bad disk? If you know that and have enough space on other disks, try moving the data on bad disk(pvmode command) to other two disks.

pvmove -n /dev/vgxx/lvolx source_pv dest_pv

Anil

[George Liu_4]

pvdisplay hangs for bad disk.

VG cannot be activated. vgchange returns error

Don't know the lvs info.

Thanks.

[Sridhar Bhaskarla]

George,

You could try activing them wiht no quorum mode. You may not be able to recover the data from the lvols that used the failed disk. But the others may be recovered.

vgchange -a y -q n vgxx

-Sri

[RAC_1]

Try actvating vgxx with qourm override.

vgchange -a y -q n vgxx.

Once activated, check which lvs reside on faulty disk. Mount those LVS with read only.

mount -F vxfs -o ro /dev/vgxx/lvolx /mount_point.

Then try copying data somewhere else.

Anil

[George Liu_4]

activate is successful, but couldn't mount. got I/O error

[Todd McDaniel_1]

You must remove that bad disk or replace it for the vg to be able to mount.

vgreduce the disk out and then you will be able to mount.

Also, mv /etc/lvmtab /tmp/lvmtab, then vgscan again to make a new one.



-----------------------------------------------------------
From the man page for vgreduce....


Force reduction of missing PVs from volume group: vg01
vgreduce -f /dev/vg01
Removes the physical volume from the volume group when it is still in
the lvmtab file, but it is currently marked as missing from the volume
group: vg01
vgreduce -l /dev/vg01

The following messages will appear after missing PVS has been removed
successfully:

PV with key 0 successfully deleted from vg /dev/vg01

Repair done, please do the following steps.....:

1. Save /etc/lvmtab to another file.

2. Remove /etc/lvmtab.

3. Use vgscan -v to recreate /etc/lvmtab.

4. NOW use vgcfgbackup(1M) to save the LVM setup.

[George Liu_4]

At this time. There is already no record for this vg in /etc/lvmtab. BTW, vgimport did something as vgscan for the file.

[Sridhar Bhaskarla]

Hi,

I haven't seen vgimport running vgscan anytime so far.

Try importing it back. It may complain that the disks are missing. Activate it using the above command and try to mount the logical volumes. If all the logical volumes have shares on the failed disk, then you will not be able to mount any of them. This is a famous draw-back of striping, which I wish you didn't have.

-Sri

[Todd McDaniel_1]

Did he say he striped them?

I read it as vg00 has 3 disks with maybe 2 mirrors...