Unable to sync a lv

UNIX AIX
1

Hi

Please look in to the below issue and advise how to fix the issue. I tried syncvg but it gives me an error. Recently we have upgraded the service pack in this lpar.

==========================================================
pmut8# lsvg -l rootvg | grep -i stale
paging00            paging     252     504     2    open/stale    N/A
==========================================================
pmut8# syncvg -l paging00
0516-1296 lresynclv: Unable to completely resynchronize volume.
        The logical volume has bad-block relocation policy turned off.
        This may have caused the command to fail.
0516-934 /usr/sbin/syncvg: Unable to synchronize logical volume paging00.
==========================================================
2

Hi,

1) you can change the "BB POLICY" to yes (relocatable) [chlv or smitty lv] and try it again (but i'm not sure if this helps) or

2) drop this paging space and create a new one if necessary.

Normaly the default paging space device (hd6) is (great) enough.
You can check/manage your paging space with lsps,chps,mkps,etc.

>lsps -a
Page Space      Physical Volume   Volume Group    Size %Used Active  Auto  Type
hd6             hdisk0            rootvg        4096MB     1   yes   yes    lv

regards

3

Does this suggest that you have an error on your disk to you? It does to me. You might be able to get away with removing the state copy of the LV and then re-mirroring.
If the suggestion from -=XrAy=- doesn't sort you out, then can you give us the output from:-

lsvg -p rootvg
lsvg -l rootvg
lslv -M hd6

Then, for each disk returned in the first command:-

lspv -l hdiskn
lspv -M hdiskn

I hope that this helps.

Robin,
Liverpool/Blackburn
UK

4

Assuming this is a physical disk - the most direct advice, given by -=XrAy=- - is to remove the paging00 logical volume.

Also simple, is to repeat the command and see if the error/warning repeats again. The key word in the message is "may". You should also review the errpt output for any specific HW related errors.

From: FAQ: errpt: Types of DISK_ERR

Rather than replace the disk immediately (maybe replacements are not possible!?) you can also try a "low-level" format and certify of the disk using the diag command in the area identified by Task Selection (Diagnostics, Advanced Diagnostics, Service Aids, etc.)

Lastly, once you identify a BB (bad block) and know which PP (physical partition) it is in, you can create a "dummy" logical volume (flag it as type lvBB using mklv -t lvBB ...) and then never assign any data to it. This will keep the AIX LVM from ever trying to use it dynamically.

Note: for this last to work you must supply some extra arguments I do not know from memory (maybe someone else does). But hsi is what manuals are for.

Note2: this assumes it is a physical volume - i.e. a real SCSI or SSA disk - directly attached to the system.