Can't do an importvg on Standby node because PV is marked as missing.
Bernie
I am running AIX 4.3.3 RML06 and HACMP 4.3.1 and am a bit of a novice with AIX
administration. Recently I added disks to a volume group on my production node
but when I tried updating the backup node, I could not import the vg because
the system claimed one of the physical volumes was marked as missing.
I did a varyonvg -b -u on my produciton node then tried importing on the backup
node using importvg -yprod03 -n hdisk12 and got an error (the same as that
below). I then tried using the "missing" disk in the volume group (hdisk73)
as the disk to import off and got the same message.
I am able to run an lqueryvg -p hdisk73 -At (again the "missing" disk) and get
the following output. From what I have been able to find on the IBM site and
Deja, I probably need to run chdev -l hdisk73 -a pv=yes, but when I told a
consultant that I did this he almost choked (though I had been advised to do
this by IBM ). Does anyone know why the chdev command with those parameters is
so dangerous?
Downtime is difficult to get so re-boots are a last resort. any ideas on this?
The following is the output from the commands I have run. I don't want to go
experimenting too much with this because it is running properly on my
production box but I would like to get it up and running on my standby.
any suggestions on this would be appreciated.
root:/home/berniem >importvg -yprod03 -n hdisk12
PV Status: hdisk12 0009334584801ee2 PVACTIVE
hdisk34 0009334584802548 PVACTIVE
hdisk67 0009334633ae0617 PVACTIVE
00093345610785cc NONAME
0516-056 varyonvg: The volume group is not varied on because a
physical volume is marked missing. Run diagnostics.
0516-780 importvg: Unable to import volume group from hdisk12.
root:/home/berniem >lqueryvg -p hdisk73 -At
Max LVs: 512
PP Size: 23
Free PPs: 2150
LV count: 22
PV count: 4
Total VGDAs: 4
Conc Allowed 0
MAX PPs per 4064
MAX PVs: 32
Quorum Setti 0
Auto Varyon 0
Conc Autovar 0
Varied on Co 0
Logical: 00093345848033a1.1 prod_0128_012 1
00093345848033a1.2 prod_0008_012 1
00093345848033a1.3 prod_0016_049 1
00093345848033a1.4 prod_0016_005 1
00093345848033a1.5 prod_0032_016 1
00093345848033a1.6 prod_0008_006 1
00093345848033a1.7 prod_0096_006 1
00093345848033a1.8 prod_0296_001 1
00093345848033a1.9 prod_0232_001 1
00093345848033a1.13 prod_0032_005 1
00093345848033a1.19 prod_0008_028 1
00093345848033a1.22 prod_0032_002 1
00093345848033a1.23 prod_1024_015 1
00093345848033a1.27 prod_0304_002 1
00093345848033a1.32 prod_0496_001 1
00093345848033a1.34 prod_0016_021 1
00093345848033a1.38 prod_0008_061 1
00093345848033a1.42 prod_1024_012 1
00093345848033a1.45 prod_0096_001 1
00093345848033a1.48 prod_0008_072 1
00093345848033a1.51 prod_0064_013 1
00093345848033a1.52 prod_0256_007 1
Physical: 00093345610785cc 1 0
0009334633ae0617 1 0
0009334584801ee2 1 0
0009334584802548 1 0
Total PPs: 3242
root:/home/berniem >importvg -yprod03 -n hdisk73
PV Status: hdisk12 0009334584801ee2 PVACTIVE
hdisk34 0009334584802548 PVACTIVE
hdisk67 0009334633ae0617 PVACTIVE
00093345610785cc NONAME
0516-056 varyonvg: The volume group is not varied on because a
physical volume is marked missing. Run diagnostics.
0516-780 importvg: Unable to import volume group from hdisk73.
Thanks
Bernie.
Jose Pina Coelho
news:7zO%6.6$6L....@nsw.nnrp.telstra.net...
> I did a varyonvg -b -u on my produciton node then tried importing on the
backup
> node using importvg -yprod03 -n hdisk12 and got an error (the same as that
> below). I then tried using the "missing" disk in the volume group
(hdisk73)
> as the disk to import off and got the same message.
>
> I am able to run an lqueryvg -p hdisk73 -At (again the "missing" disk) and
get
> the following output. From what I have been able to find on the IBM site
and
> Deja, I probably need to run chdev -l hdisk73 -a pv=yes,
> but when I told a consultant that I did this he almost choked
> (though I had been advised to do this by IBM ).
Did you tell them that you had HACMP and that the PV was already part of a
VG ?
> Does anyone know why the chdev command with those parameters is
> so dangerous?
Because it will create a new PV id for that disk, essentialy breaking the
association
between the disk and the VG. A true and complete mess.
> Downtime is difficult to get so re-boots are a last resort. any ideas on
this?
Yes, you don't need downtime.
> The following is the output from the commands I have run. I don't want to
go
> experimenting too much with this because it is running properly on my
> production box but I would like to get it up and running on my standby.
Good option.
After the varyonvg -b -n on the production node, run a cfgmgr -l scsix/ssax
(whatever the card that communicates with the disk).
That should update your machine to recognize the new disks. Then do the
importvg
Get the HACMP manuals. Read them until you know them by heart.
--
Doing AIX support was the most monty-pythonesque
activity available at the time.
Nick Buckley
Agreed, the relationship between ODM and VGDA is critical, but there is a
way out of the mess.
Beware though, it's a hack that I have used and as such is AT YOUR OWN RISK.
If hdisk73 is your "bogey" disk, then:
i) lquerypv -H /dev/hdisk73 - this is the new PVID written to the disk by
chdev - you need this.
ii) odmget -q"name=hdisk73 and attribute=pvid" CuAt > /tmp/foo - capture the
ODM entry
iii) odmdelete -o CuAt -q"name=hdisk73 and attribute=pvid" - delete this
entry
iv) vi /tmp/foo and replace the PVID ( including the 16 trailing zeroes )
with the output of i) - you must include 16 trailing zeroes to make a 32
digit number.
v) odmadd /tmp/foo - replace entry into ODM with the correct PVID
You have now restored the 1:1 relationship between PVID in VGDA and PVID in
ODM.
You should now be able to run importvg -yprod03 -n hdisk12 without any
errors.
Regards,
Nick Buckley,
Senior AIX Administrator,
NCM,
Cardiff,
United Kingdom
e-mail: nicholas...@ncmgroup.com
(Please note: all newsgroup offerings are made in a personal capacity and in
no way are the responsibility of my employer)
p.s. Did you know that "AIX Operating System" is an anagram of "Mega sexy
partitions" !!!
Jose Pina Coelho <bo...@clix.pt> wrote in message
news:v9P%6.10501$MK1.28...@newsserver.ip.pt...
> "Bernie" <bernie.mil...@wesley.com.au> wrote in message
> news:7zO%6.6$6L....@nsw.nnrp.telstra.net...
> >