vgchange fails

[Enrico Venturi]

Colleagues,
I receive this error when I issue the command:
#vgchange -a e /dev/vg_161

output:
vgchange: Warning: Couldn't attach to the volume group physical volume "/dev/dsk/c4t8d0":
Cross-device link
vgchange: Activation mode requested for the volume group "/dev/vg_161" conflicts with configured mode.
vgchange: Warning: couldn't query physical volume "/dev/dsk/c4t8d0":
The specified path does not correspond to physical volume attached to
this volume group
vgchange: Warning: couldn't query all of the physical volumes.

The ioscan -fnCdisk command provides this output:

lass I H/W Path Driver S/W State H/W Type Description
============================================================================
disk 0 0/0/1/1.0.0 sdisk CLAIMED DEVICE SEAGATE ST39204LC
/dev/dsk/c1t0d0 /dev/rdsk/c1t0d0
disk 1 0/0/2/0.0.0 sdisk CLAIMED DEVICE SEAGATE ST39204LC
/dev/dsk/c2t0d0 /dev/rdsk/c2t0d0
disk 2 0/0/2/1.2.0 sdisk CLAIMED DEVICE HP DVD-ROM 305
/dev/dsk/c3t2d0 /dev/rdsk/c3t2d0
disk 3 0/7/0/0/4/0.0.0 sdisk CLAIMED DEVICE SEAGATE ST318404LC
/dev/dsk/c4t0d0 /dev/rdsk/c4t0d0
disk 4 0/7/0/0/4/0.1.0 sdisk CLAIMED DEVICE SEAGATE ST318404LC
/dev/dsk/c4t1d0 /dev/rdsk/c4t1d0
disk 5 0/7/0/0/4/0.8.0 sdisk CLAIMED DEVICE SEAGATE ST318404LC
/dev/dsk/c4t8d0 /dev/rdsk/c4t8d0
disk 6 0/7/0/0/4/0.9.0 sdisk CLAIMED DEVICE SEAGATE ST318404LC
/dev/dsk/c4t9d0 /dev/rdsk/c4t9d0
disk 7 0/7/0/0/4/1.0.0 sdisk CLAIMED DEVICE SEAGATE ST318404LC
/dev/dsk/c5t0d0 /dev/rdsk/c5t0d0
disk 8 0/7/0/0/4/1.1.0 sdisk CLAIMED DEVICE SEAGATE ST318404LC
/dev/dsk/c5t1d0 /dev/rdsk/c5t1d0
disk 9 0/7/0/0/4/1.2.0 sdisk CLAIMED DEVICE SEAGATE ST318404LC
/dev/dsk/c5t2d0 /dev/rdsk/c5t2d0
disk 10 0/7/0/0/4/1.8.0 sdisk CLAIMED DEVICE SEAGATE ST318404LC
/dev/dsk/c5t8d0 /dev/rdsk/c5t8d0
disk 11 0/7/0/0/4/1.9.0 sdisk CLAIMED DEVICE SEAGATE ST318404LC
/dev/dsk/c5t9d0 /dev/rdsk/c5t9d0


Any suggestions?

Thanks to everybody

Enrico

[Christopher McCray_1]

Hello,

Has this disk by chance been replaced recently?

The last time we got a cross-linked device error is when our FE replaced the wrong disk.

Have you tried the vgsync command?

Also, try doing a vgchange -c y then issuing a vgchange -a e

Can you post the vgdisplay output?

Hope this helps

Chris

[Enrico Venturi]

agnel,sys,root # vgdisplay
--- Volume groups ---
VG Name /dev/vg00
VG Write Access read/write
VG Status available
Max LV 255
Cur LV 7
Open LV 7
Max PV 16
Cur PV 2
Act PV 2
Max PE per PV 2500
VGDA 4
PE Size (Mbytes) 4
Total PE 4338
Alloc PE 4330
Free PE 8
Total PVG 0
Total Spare PVs 0
Total Spare PVs in use 0

vgdisplay: Volume group not activated.
vgdisplay: Cannot display volume group "/dev/vg_161".

[James R. Ferguson]

Hi:

Did you previously mark the volume group as part of a cluster?

# vgchange -c y ...

Regards!

...JRF...

[Enrico Venturi]

agnel,sys,root # vgchange -c y /dev/vg_161
vgchange: Warning: Couldn't attach to the volume group physical volume "/dev/dsk/c4t8d0":
Cross-device link
vgchange: Could not attach to volume group "/dev/vg_161" all of
its physical volumes.
vgchange: Warning: couldn't query physical volume "/dev/dsk/c4t8d0":
The specified path does not correspond to physical volume attached to
this volume group
vgchange: Warning: couldn't query all of the physical volumes.
agnel,sys,root # vgsync
Usage: vgsync VolumeGroupName...
More arguments required.
agnel,sys,root # vgsync /dev/vg_161
vgsync: Couldn't query logical volume "/dev/vg_161/lvol1":
Volume group not activated.
***********************
vgsync: Couldn't query the list of logical volumes.
vgsync: Couldn't access the list of logical volumes for volume group "/dev/vg_161".
vgsync: Couldn't resynchronize volume group "/dev/vg_161".

[MANOJ SRIVASTAVA]

Hi Enrico


If I assume that c4t8d0 is a part of the vg than you can do the following .

vgchange -a n /dev/vg_161
vgcfgrestore /dev/vg_161 /dev/rdsk/c4t8d0
vgchange -a y /dev/vg_161

and you should be good to go.



Manoj Srivastava

[Jeff Schussele]

Hi Enrico,

I'm sorry to say, but this is definitely not good.

1) Cross-device link usually indicates that this PV has been included in another VG - possibly on another system even. If this is a shared array, verify this PV is NOT in another VG. If it is & it shouldn't be - remove it from that unnecessary VG

2)Activation mode generally means you need to make the VG cluster-aware:
vgchange -c y /dev/vg_161

3) Couldn't query means this system does not think c4t8d0 belongs to vg_161. Check the /etc/lvmtab to see what it thinks vg_161 should contain & if it is incorrect, mv or delete it & and run vgscan -v to recreate a new one w/the correct info.

4) Couldn't query all says that again lvmtab states that vg_161 contains more/different PVs than what the system actually sees. Again this could point to lvmtab being bad or corrupt. I'd say definitely rebuild lvmtab.

HTH,
Jeff

[Victor_5]

From doc A4364386:

Check your /etc/lvmtab to see if vgextend /dev/vg161 /dev/dsk/c4d8d0 is already recorded.

This problem usually occurs because two VGs claim ownership for the same disk.

If you run:

strings /etc/lvmtab

you should see the VGs both claim the same disk (i.e., both
show /dev/dsk/c4t8d0).

Follow these steps to correct the situation:

1. mv /etc/lvmtab /etc/lmvtab.rc

2. vgscan -v

3. strings /etc/lvmtab # make sure that each volume group has a unique drive.

You should now be able to run vgchange -a y /dev.

[Enrico Venturi]

As far as I know, if the PV belongs to a VG on the system A, this isn't known by the system B which wants use the same PV on another local VG.
I mean, two systems can separately create two VG on two machines each one use the same PV.
This is an error but isn't recognised as crossed link

[Christopher McCray_1]

Hello again,

Let me see if I understand this...

The output of strings /etc/lvmtab shows an entry for the pv (hopefully in the same vg because it is apparent to me that this is a shared disk for a package).

Look at your /etc/lvmrc file. The AUTO_ACTIVATE_VG variable should be set to 0. If it isn't, then both nodes in your cluster will try to activate the vg and cause you problems.

Hope this helps

Chris

[Christopher McCray_1]

An addition

If you have a PV belonging to a VG on nodeA, then you cannot make it belong to a VG on nodeB. There is already LVM information pertaining to nodeA and the VG on that node.


Chris

[Enrico Venturi]

Collegues,
we found the solution.
The disk is shared between two nodes, as I said.
One of them has created a volume group on it, therefore the old volume group prent on the disk has been overwritten.
I was not aware of that because the disk is named in a different way on the two nodes, being that one of them is a L1000 server and the other is a N4000 server, so it's difficult to recognize the disks ....

Thanks to everubody for answering.
Enrico

[harry d brown jr]

Enrico,

This member has assigned points to 45 of 116 responses to his/her questions.

http://forums.itrc.hp.com/cm/TopSolutions/1,,BR62571!1!questions,00.html

live free or die
harry

[Enrico Venturi]

So you mean ...??
I had to assign more points to my questions? :-)