Need to re-mirror

[N.D]

Our root disk failed so we had this replace, luckily it was mirrored and we were able to boot from the mirrored. I nedd to now re-configure the original disk. Currently when doing an lvdisplay the replaced disk is showing as ???. How do I re-configure ?


--- Logical extents ---
LE PV1 PE1 Status 1 PV2 PE2 Status 2
0000 ??? 0000 stale /dev/dsk/c2t15d0 0000 current
0001 ??? 0001 current /dev/dsk/c2t15d0 0001 stale
0002 ??? 0002 stale /dev/dsk/c2t15d0 0002 current
0003 ??? 0003 stale /dev/dsk/c2t15d0 0003 current

[Zigor Buruaga]

Hi,

Attached the instructions to replace a root mirrored disk.

Kind regards,
Zigor

[Bernhard Mueller · Accepted solution]

Hi,

I am a bit worried about the output you send regarding logical extend 0001: it shows "current" on the replaced disk and "stale" on the good one...

Basically it would go like:
strings /etc/lvmtab (find out what is the replaced disks device file, supposed it is c3t3d0 belonging to vg00... then:
vgcfgrestore -n vg00 /dev/rdsk/c3t3d0
vgchange -a y vg00
vgsync vg00

Good luck
Bernhard

[A. Clay Stephenson]

Let's assume that your replaced disk is c0t6d0.

1) vgcfgrestore -n /dev/vg00 /dev/rdsk/c0t6d0
2) vgchange -a y /dev/vg00
3) mkboot /dev/rdsk/c0t6d0
4) mkboot -a "hpux -lq (;0)/stand/vmunix" /dev/rdsk/c0t6d0
5) lvlnboot -R
6) vgsync /dev/vg00

You are done. I would then shutdown and verify that you can boot from the restored primary now.

[N.D]

got the following reply from the 'vgrestore' command

# vgcfgrestore -n /dev/vg00 /dev/rdsk/c2t6d0
vgcfgrestore: Couldn't query physical volume "/dev/dsk/c2t6d0":
The specified path does not correspond to physical volume attached to
this volume group
Volume Group configuration has been restored to /dev/rdsk/c2t6d0

[Bernhard Mueller]

Try vgdisplay -v vg00 and see if c2t6d0 is listed as available.

[N.D]

Looks like its syncing fine .... so will the new replaced root disk automatically mirror data from the mirror disk.

[Bryan D. Quinn]

Hey Missy,

I had a failed boot drive on one of my N-4000's two nights ago. I got the same errors you are seeing. I have attached the procedure that I wrote up to handle it and get my system back up.

I must say, I don't understand why you have a discrepency in the stale/current state on extent 0001.

Hope this helps you!
-Bryan

[Bernhard Mueller]

check pvdisplay /dev/dsk/c2t6d0 | grep Stale

[Bernhard Mueller]

do not forget the mkboot and lvlnboot commands that Clay mentioned, *after* that
check lvlnboot -v (that c2t6d0 is a "Boot Disk") then check setboot to see if it is primary or alternate boot path.

*If* you can, try to boot from the failed disk to verify, if not take note to get a downtime for reboot. Halt at BCH and enter
bo pri
or
bo alt
(whichever it is)

Regards,
Bernhard