Mirror disk failure

[markjr2006]

Houston.. I think there is a problem..

See out put of lvdisplay -v /dev

--- Logical volumes ---
LV Name /dev/VG01/lvola
VG Name /dev/VG01
LV Permission read/write
LV Status available/stale
Mirror copies 1
Consistency Recovery MWC
Schedule parallel
LV Size (Mbytes) 8676
Current LE 2169
Allocated PE 4338
Stripes 0
Stripe Size (Kbytes) 0
Bad block on
Allocation strict
IO Timeout (Seconds) default

--- Distribution of logical volume ---
PV Name LE on PV PE on PV
/dev/dsk/c0t14d0 2169 2169

--- Logical extents ---
LE PV1 PE1 Status 1 PV2 PE2 Status 2
0000 /dev/dsk/c0t14d0 0000 current ??? 0000 stale
0001 /dev/dsk/c0t14d0 0001 current ??? 0001 stale
0002 /dev/dsk/c0t14d0 0002 current ??? 0002 stale
0003 /dev/dsk/c0t14d0 0003 current ??? 0003 stale
0004 /dev/dsk/c0t14d0 0004 current ??? 0004 stale
0005 /dev/dsk/c0t14d0 0005 current ??? 0005 stale
0006 /dev/dsk/c0t14d0 0006 current ??? 0006 stale
0007 /dev/dsk/c0t14d0 0007 current ??? 0007 stale
0008 /dev/dsk/c0t14d0 0008 current ??? 0008 stale
0009 /dev/dsk/c0t14d0 0009 current ??? 0009 stale
0010 /dev/dsk/c0t14d0 0010 current ??? 0010 stale
0011 /dev/dsk/c0t14d0 0011 current ??? 0011 stale
0012 /dev/dsk/c0t14d0 0012 current ??? 0012 stale
0013 /dev/dsk/c0t14d0 0013 current ??? 0013 stale
0014 /dev/dsk/c0t14d0 0014 current ??? 0014 stale
0015 /dev/dsk/c0t14d0 0015 current ??? 0015 stale
0016 /dev/dsk/c0t14d0 0016 current ??? 0016 stale
0017 /dev/dsk/c0t14d0 0017 current ??? 0017 stale
0018 /dev/dsk/c0t14d0 0018 current ??? 0018 stale
0019 /dev/dsk/c0t14d0 0019 current ??? 0019 stale
0020 /dev/dsk/c0t14d0 0020 current ??? 0020 stale
0021 /dev/dsk/c0t14d0 0021 current ??? 0021 stale
0022 /dev/dsk/c0t14d0 0022 current ??? 0022 stale

I think the disk which has failed is c0t15d0

How do I remove the failed disk from the mirror until I get a new one?

Many thanks for any help!!

[Antonio Cardoso_1]

Hi,

read this first,
http://docs.hp.com/en/5991-1236/When_Good_Disks_Go_Bad.pdf

Check disk status using
ioscan -fnC disk

If hotplugable disks, you can safely remove bad one without stopping your server. Check your server's characteristics for that.

[Shahul]

Hi,

I think you have just lost the second hard disk of this mirror. To reduce the mirror you can use lvreduce with -m option. To remove mirror from a missing/failed PV you can use lvreduce with -m & -k option.

See man lvreduce for further details.

You can also remove missing hard disk from a vg by vgreduce with -f option.

Hope this helps.

Best of luck
Shahul

[Prashant Zanwar_4]

If disk is not bad you can do following:

vgchange -a y

It will try to resynchronize your VG and get rid of stale extents..

dmesg will tell if disk is bad

dd if=/dev/rdsk/ of=/dev/null bs=1024 - should not complete successfully if disk were bad..

gud luk
Prashant

[Sivakumar TS]

Hi,

As the disk is not available, then the lvreduce command will fail with an error that it could not query the physical volume.

You can still remove the mirror copy, but you must specify the physical volume key rather than the name. You can get the key using lvdisplay with the â k option.

Regards,

Siva.

[Sivakumar TS]

Hi,

use...

# lvdisplay -v â k /dev/VG01/lvola

note down the key of the disk which shows "stale"

then

#lvreduce -m 0 -A n â k /dev/VG01/lvola

Now you are done !

Regards,

Siva.

[markjr2006]

lvreduce failed as noted.

Tried lvdisplay -v -k to obtain key and I'm not sure if this looks right?

output of lvdisplay -v -k /dev/VG01/lvola

0000 0 0000 current 1 0000 stale
0001 0 0001 current 1 0001 stale
0002 0 0002 current 1 0002 stale
0003 0 0003 current 1 0003 stale
0004 0 0004 current 1 0004 stale
0005 0 0005 current 1 0005 stale
0006 0 0006 current 1 0006 stale
0007 0 0007 current 1 0007 stale
0008 0 0008 current 1 0008 stale
0009 0 0009 current 1 0009 stale
0010 0 0010 current 1 0010 stale
0011 0 0011 current 1 0011 stale
0012 0 0012 current 1 0012 stale
0013 0 0013 current 1 0013 stale
0014 0 0014 current 1 0014 stale
0015 0 0015 current 1 0015 stale
0016 0 0016 current 1 0016 stale
0017 0 0017 current 1 0017 stale
0018 0 0018 current 1 0018 stale
0019 0 0019 current 1 0019 stale
0020 0 0020 current 1 0020 stale
0021 0 0021 current 1 0021 stale
0022 0 0022 current 1 0022 stale


**Simakumar**
So I just wanted to make sure that I'm thinking correctly.. The key then would be "1" for the failed disk?

Thanks

[markjr2006]

lvreduce failed as noted.

Tried lvdisplay -v -k to obtain key and I'm not sure if this looks right?

output of lvdisplay -v -k /dev/VG01/lvola

0000 0 0000 current 1 0000 stale
0001 0 0001 current 1 0001 stale
0002 0 0002 current 1 0002 stale
0003 0 0003 current 1 0003 stale
0004 0 0004 current 1 0004 stale
0005 0 0005 current 1 0005 stale
0006 0 0006 current 1 0006 stale
0007 0 0007 current 1 0007 stale
0008 0 0008 current 1 0008 stale
0009 0 0009 current 1 0009 stale
0010 0 0010 current 1 0010 stale
0011 0 0011 current 1 0011 stale
0012 0 0012 current 1 0012 stale
0013 0 0013 current 1 0013 stale
0014 0 0014 current 1 0014 stale
0015 0 0015 current 1 0015 stale
0016 0 0016 current 1 0016 stale
0017 0 0017 current 1 0017 stale
0018 0 0018 current 1 0018 stale
0019 0 0019 current 1 0019 stale
0020 0 0020 current 1 0020 stale
0021 0 0021 current 1 0021 stale
0022 0 0022 current 1 0022 stale


**Sivakumar**
So I just wanted to make sure that I'm thinking correctly.. The key then would be "1" for the failed disk?

Thanks

[markjr2006]

Furthermore...


vgdisplay -v VG01
vgdisplay: Warning: couldn't query physical volume "/dev/dsk/c0t15d0":
The specified path does not correspond to physical volume attached to
this volume group
--- Logical volumes ---
LV Name /dev/VG01/lvola
LV Status available/stale
LV Size (Mbytes) 8676
Current LE 2169
Allocated PE 4338
Used PV 1


vgdisplay: Warning: couldn't query physical volume "/dev/dsk/c0t15d0":
The specified path does not correspond to physical volume attached to
this volume group
vgdisplay: Warning: couldn't query all of the physical volumes.
--- Physical volumes ---
PV Name /dev/dsk/c0t14d0
PV Status available
Total PE 2170
Free PE 1

Output of ioscan -fCdisk
Class I H/W Path Driver S/W State H/W Type Description
=====================================================================
disk 2 8/0.1.0 sdisk CLAIMED DEVICE SEAGATE ST318436LC
disk 3 8/0.2.0 sdisk CLAIMED DEVICE SEAGATE ST318436LC
disk 6 8/0.14.0 sdisk CLAIMED DEVICE SEAGATE ST39236LC
disk 0 8/16/5.2.0 sdisk CLAIMED DEVICE TOSHIBA CD-ROM XM-6201A
disk 1 8/16/5.6.0 sdisk CLAIMED DEVICE SEAGATE ST15150N

[Prashant Zanwar_4]

You are on the device which is bad.. do you have backup of LV you are working with, if not back it up now...

You are loosing c0t15d0 and after lvreduce you are on failing drive..

Now only bet is lvremove..

If I were you, i would have tried extending LV back,

lvextend -m 1 /dev/vgname /dev/dsk/ ..

then do lvreduce

lvreduce -A n -m 0 -k /dev/vgname /dev/dsk/c0t15d0 (you have to tell in command what disk you want to get rid off..

also

vgreduce vgname /dev/dsk/c0t15d0

once lvreduce is done.. replace disk

after disk is replaced.. do the follwoing

pvcreate /dev/rdsk/c0t15d0
vgextend /dev/vgname /dev/dsk/c0t15d0
lvextend -m 1 /dev/vgname /dev/dsk/c0t15d0

If required restore contents of LV/FS..

You should be good now..

I am not sure if you will be able to backup data..

Please be cautious of every step you are taking

Good luck
Regards
Prashant

[Prashant Zanwar_4]

message for shahul:
Hey just add me in ur aol / yahoo:
Sorry for posting here
aol :- pzanwar
yahoo/hot :- pczanwar

Thx
Prashant

[Prashant Zanwar_4]

Remove the disk after vgreduce.. (My mistake earlier tell after lvreduce)..

Tx
Prashant