Very basic SG issue - initial install

[Steven Buschman_3]

Serviceguard novice, been battling this for days. Non-production, just need to develop an agent for 11.16. I've downloaded countless documents from hp.com/hpux/ha and searched many posts. I cannot find a doc to do the initial setup, I can't seem to get the quorum (lock) disk going so I can't get to the point of establishing the cluster where the managing documentation starts. I've tried many /usr/sbin/cm* commands, man pages, and SAM but everything I've tried picks up later than where I am. I feel I'm cluster knowledgabe otherwise. 2 node (3440) and a 4 disk scsi array. Yes the controllers are set to 6 and 7 already, array drives are 8,10,12,14 - "Simple" 8-)
Can anyone tell me where to go (be nice...)
Thanks,
Steven

[A. Clay Stephenson · Accepted solution]

Rather than having you describe your setup -- and maybe lying (unintentionally), let's ask the boxes themselves so that we can all start out on the same page.

Run this command and post the results:
cmquerycl -v -n nodeA -n nodeB

[Steven Buschman_3]

Here's the output from cmquerycl....
Please not I have been adding and removing LV's in attempts to get things in the right order. At this moment I deleted the lock disk I had defined out of LVM (called /dev/quorum, c3t8d0 was used).
I saw documentation where the lock disk needs to be created and under LVM (not CVM or VxVM) but at this moment it was removed.
Thanks!
Steven

# cmquerycl -v -n hpmcsg1 -n hpmcsg2

Begin checking the nodes...
Gathering configuration information ........ Done

Node Names: hpmcsg1
hpmcsg2

Bridged networks:

IP subnets:

IPv4:

10.80.136.0 lan0 (hpmcsg1)
lan0 (hpmcsg2)

192.168.253.0 lan1 (hpmcsg1)
lan1 (hpmcsg2)

192.168.252.0 lan2 (hpmcsg1)
lan2 (hpmcsg2)

IPv6:

Possible Heartbeat IPs:

10.80.136.0 10.80.137.171 (hpmcsg1)
10.80.137.172 (hpmcsg2)

192.168.253.0 192.168.253.1 (hpmcsg1)
192.168.253.2 (hpmcsg2)

192.168.252.0 192.168.252.1 (hpmcsg1)
192.168.252.2 (hpmcsg2)

Possible Cluster Lock Devices:

LVM volume groups:

/dev/vg00 hpmcsg1

/dev/vg00 hpmcsg2

LVM physical volumes:

/dev/vg00
/dev/dsk/c2t0d0 0/1/1/0.0.0 hpmcsg1

/dev/vg00
/dev/dsk/c2t0d0 0/1/1/0.0.0 hpmcsg2

LVM logical volumes:

Volume groups on hpmcsg1:
/dev/vg00/lvol1 FS MOUNTED /stand
/dev/vg00/lvol2
/dev/vg00/lvol3 FS MOUNTED /
/dev/vg00/lvol4 FS MOUNTED /exthome
/dev/vg00/lvol5 FS MOUNTED /opt
/dev/vg00/lvol6 FS MOUNTED /tmp
/dev/vg00/lvol7 FS MOUNTED /usr
/dev/vg00/lvol8 FS MOUNTED /var

Volume groups on hpmcsg2:
/dev/vg00/lvol1 FS MOUNTED /stand
/dev/vg00/lvol2
/dev/vg00/lvol3 FS MOUNTED /
/dev/vg00/lvol4 FS MOUNTED /exthome
/dev/vg00/lvol5 FS MOUNTED /opt
/dev/vg00/lvol6 FS MOUNTED /tmp
/dev/vg00/lvol7 FS MOUNTED /usr
/dev/vg00/lvol8 FS MOUNTED /var

[melvyn burnard]

As this is a two node cluster, you MUST have some form of cluster locking.
Whether you set up a separate Quorum Server, or you use a lock disc, is up to you.#
If you wish to use the disc, you must first set up a VG within LVM, using at least one PV that is seen by both nodes.

Once that is done, ensure the VG is active on on of the nodes, then do cmquerycl -v -C /etc/cmcluster/ -n node1 -n node2

You should then see the cluster lock disc in the ascii file if all was set correctly.

When you then run cmapplyconf, with that VG active, it will write the cluster lock bits to the disc, and set it into the binary.

This is all documented in the Managing Serviceguard manual at:
http://docs.hp.com/hpux/ha

[Sanjay_6]

Hi,

From your post, it looks as if you only have vg00 configured on the two nodes. vg00 is always local to individual nodes.

You need to configure atleast one more vg and share it between the two nodes. A disk which is part of this shared vg should be configured as the lock disk (you need one, since this is a 2 node cluster).

Hope this helps.

Regds