Mirroring VG00 gone bad

[Darren Etheridge_2]

I am trying to mirror our 2 root drives VG00 -
/dev/dsk/c0t6d0 ==> /dev/dsk/c3t6d0

when trying to mirror the ldevs i get the following error:
lvextend: Couldn't re-synchronize stale partitions of the logical volume:
Device offline/Powerfailed

lvextend: Couldn't resynchronize logical volume "/dev/vg00/lvol5


Bad disk? If so, how can I be sure it's not c0t6d0? thanks in advance for your help!

[James R. Ferguson · Accepted solution]

Hi Darren:

> Bad disk? If so, how can I be sure it's not c0t6d0?

Try:

# dd if=/dev/rdsk/c0t6d0 of=/dev/null bs=1024k count=10

Regards!

...JRF...

[smatador]

Hi,

Look at the lvdisplay /dev/vg00/lvol5 | grep stale
and
pvdisplay -v /dev/dsk/c0t6d0
pvdisplay -v /dev/dsk/c3t6d0
If you have a pv state unavailable or some stale status that mean you have a bad disk.
HTH

[AVV]

Hi,

Try with

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

check with the pvdisplay

and do a dd to whole disk .

[Darren Etheridge_2]

Both Seem to come back ok


dd if=/dev/rdsk/c0t6d0 of=/dev/null bs=1024k count=10

10+0 records in
10+0 records out


dd if=/dev/rdsk/c3t6d0 of=/dev/null bs=1024k count=10

10+0 records in
10+0 records out

[Darren Etheridge_2]

below are the results of the pvdisplay. I am interperting my c3t6d0 is bad. Correct?

[TAL:/]pvdisplay -v /dev/dsk/c0t6d0 |grep stale


[TAL:/]pvdisplay -v /dev/dsk/c3t6d0 |grep stale
00091 stale /dev/vg00/lvol2 00059
00152 stale /dev/vg00/lvol2 00120
00162 stale /dev/vg00/lvol2 00130
00165 stale /dev/vg00/lvol2 00133
00168 stale /dev/vg00/lvol2 00136
00201 stale /dev/vg00/lvol2 00169
00283 stale /dev/vg00/lvol2 00251
00310 stale /dev/vg00/lvol2 00278
00314 stale /dev/vg00/lvol2 00282
00318 stale /dev/vg00/lvol2 00286
00334 stale /dev/vg00/lvol2 00302
00346 stale /dev/vg00/lvol2 00314
00351 stale /dev/vg00/lvol2 00319
00408 stale /dev/vg00/lvol2 00376

[Darren Etheridge_2]

how do i dd a whole disk?

[Patrick Wallek]

>>how do i dd a whole disk?

Remove the "count=10" from the dd command line.

[Darren Etheridge_2]

That's what I thought... I just wanted to check b4 messing up something :)

[James R. Ferguson]

Hi (again):

> That's what I thought... I just wanted to check b4 messing up something :)

In this case all we are doing is reading the disk designated by the 'if' argument. It would be a catastrophic mess only if you exchanged the 'if' and 'of' devices in my example :-)

I chose a 'count=10' to generate a quick look. A full 'dd' even given the raw device and the optimal block size will take considerable time. Be patient.

Regards!

...JRF...

[Darren Etheridge_2]

Thanks everyone for your help. Drive was ordered and I just got it in... I am rebuilding the mirrors now... Have a great week!

[Johnson Punniyalingam]

Hi Darren,

You can also refer to HP offical document

"Good disk Gone Bad".

http://docs.hp.com/en/5991-1236/When_Good_Disks_Go_Bad_WP.pdf

Thanks,
Johnson

[R.K. #]

Hello Darren,

Just wanted to know if you did these:

Status of the disk?
# pvdisplay /dev/dsk/c0t6d0 | grep "PV Status"

Try enabling the disk for IOs (if it is disabled)
# pvchange -a y /dev/dsk/c0t6d0

Regds..