vgimport error MC/SG

[hpuxsa]

I am trying to import a vg on to the standby node in a MC/SG cluster and i am getting the following error. How do I import it on both the PV's. The disks are visible on both the paths.

vgimport -p -v -m /tmp/vg01.map vg01 /dev/dsk/c25t0d1 /dev/dsk/c29t0d1
Beginning the import process on Volume Group "vg01".
vgimport: Warning: Volume Group belongs to different CPU ID.
Can not determine if Volume Group is in use on another system. Continuing.
vgimport: Warning: Volume Group contains "1" PVs, "2" specified. Continuing.
Logical volume "/dev/vg01/lvol1" has been successfully created
with lv number 1.
Volume group "/dev/vg01" has been successfully created.
Warning: A backup of this volume group may not exist on this machine.
Please remember to take a backup using the vgcfgbackup command after activating
the volume group.

Appreciate your help.
Thanks in advance.

[Massimo Bianchi]

The sintax is ccorrect, you are only seeing a warning because the VG was created on the oyher server.


To do the real import, just tear away the "-p"
vgimport -p -v -m /tmp/vg01.map vg01 /dev/dsk/c25t0d1 /dev/dsk/c29t0d1

I say it a second time: they are NORMAL warnings, they are not errors.

Massimo

[Tim D Fulford]

Hi

I think the command was successful. see the line at the end

> Volume group "/dev/vg01" has been successfully created. <

As Massimo says strip out the -p option to do the import. I assume that the VG consists of 1 disk with 2 alternate paths (hence the warning can be ignored)

Regards

Tim

[Rita C Workman]

Before I vgimport I always get rid of what is out in lvmtab first:

vgexport /dev/vgname
mkdir /dev/vgname
mknod /dev/vgname/group c 64 0x_ _ 0000

Then I do my import ( I tend to use the -s option, but that's just me.. ):

vgimport -vs -m /etc/lvmconf/vgname.map /dev/vgname

I do not tend to extend the statement with the disk device info (c_t_d_) as you noted above.

Just a thought,
Rita

[Kent Ostby]

A document that might help you with this process is:

http://www2.itrc.hp.com/service/cki/search.do?category=c0&mode=id&searchString=UXSGKBAN00000037&searchCrit=allwords&docType=Security&docType=Patch&docType=EngineerNotes&docType=BugReports&docType=Hardware&docType=ReferenceMaterials&docType=ThirdParty&search.x=30&search.y=5

It is a certified document related to distributing the VGs between SG nodes.

Best regards,

Kent M. Ostby

[Sanjay_6]

The way we do this is,

on the 1st node,

vgchange -a n /dev/vgname
vgexport -p -s -m /tmp/vgname.map /dev/vgname

rcp the /tmp/vgname.map file to the second node, then do a vgimport on the other node.

mkdir /dev/vgname
mknod /dev/vgname/group c 64 0x0n0000
vgimport -s -m /tmp/vgname.map /dev/vgname
vgchange -a n /dev/vgname
vgchange -a y /dev/vgname

If this was a cluster aware vg, use vgchange -a e /dev/vgname to activate the vg in exclusive mode.

mount the filesystems and check and see if the vg is correctly imported or not.

Hope this helps.

Regds

[hpuxsa]

i did a vgimport on the standby node only after exporting the vg. There was no reference to the vg in the lvmtab file of the standby node. But when i try to vgimport it gives me the Warning:Volume Group contains "1" PVs, "2" specified. Where does the standby node get the info that the vg has only one PV.
I tried importing the vg ignoring the warning and it updates the lvmtab file with Alternate PV path also.

I was trying to recreate the cluster config file "cmquerycl -v -n node1 -n node2 -C filename". The config file created is missing the values for
FIRST_CLUSTER_LOCK_VG
FIRST_CLUSTER_LOCK_PV
VOLUME_GROUP

[Sandro Schaer_2]

check the vg on your main server by using 'vgdisplay -v /dev/vg01'. what does it show under 'Cur PV' ? This is the number of pv's on the main server.

I do the vgexport/import usually as follows :

on the main server :
vgexport -p -s -v -m /tmp/vg01.map /dev/vg01

on the standby host(s) :
- vgexport /dev/vg01
- ioscan -fnCdisk
- insf -e
- get vg01.map from main server, save into /tmp
- mkdir /dev/vg01
- mknod /dev/vg01/group c 64 0xnn0000 (same as on main server)
- vgimport -s -v -m /tmp/vg01.map /dev/vg01


Never had any difficulties with this procedure. Running a 8-node cluster with 28 packages and almost 60 vg's spread across various disk systems.


hope this helps


have fun
Sandro

[Tim D Fulford]

Hi

>But when i try to vgimport it gives me the
>Warning:Volume Group contains "1" PVs, "2"
>specified. Where does the standby node get >the info that the vg has only one PV.

The warning is because you have alternate paths set up. LVM stores this information on the header & footer of the disks themselves & they KNOW that there is only 1 physical disk (or LUN I assume), & when you do the vgimport you are specifying two disks/paths (hence the eWarning message). At that point the programs obviously do not know one path is an alternate PV link.

IF you were to do
vgexporet -s -m /tmp/vg01.map -p -v vg01
&
vgimport -s -m /tmp/vg01.map -v vg01

you would not see this message, as the vgimport goes off & scans the hardware & finds all the disks with the correct VGID.

There really is nothing to worry about, it is an warning that you can ignore.

Regards

Tim