LOCK VG cluster aware and cmapplyconf

[Francesco Campalastri]

I suppose this is a famous question:

My cmcheckconf fails due to Lock VG not submitted to cluster.

But vgchange -c y fails due to cluster undefined.

How to get it?

[Rajeev Shukla]

Hi,
Is this a new cluster you are configuring?

And the disk and VG you have choosen for Cluster Lock should be sharable(seen by every node in the cluster). Otherwise choose a cluster disk which is visible to all.

How many nodes do you have if your cluster? coz cluster Lock can be avoided if you have a cluster of 4 or mode nodes.

Rajeev

[Sridhar Bhaskarla]

Hi,

When you apply the configuration for the first time, you will need to have all the volume groups part of the cluster should be "active" on the server where you are running cmapplyconf. Do this.

#vgchange -a y vg01
#vgchange -a y vg02
etc.,
#cmapplyconf -C /etc/cmcluster/clusterconf.ascii

"cmapplyconf" will take care of making the VGs clusteraware.

-Sri

[Francesco Campalastri]

Thanks for answers.

Cluster is a new 2 nose one. Cluster Lock Disk is shared, I mean both nodes can activate the cluster logk vg with vgchange -a y.

I know I have to activate the vg before making the apply but the problem is: cmapplyconf fails!

I have Sevice Guard 11.14 with shared LUN on an Eva Storage and Secure path to access then with load balancing.

Any idea?

[Rajeev Shukla]

What are errors you are getting when using cmapplyconf also see what errors /var/adm/syslog/syslog is reporting.

Also make sure that the both the machines should be able to rlogin or run a remote shell before you do that.
Have the entries in /etc/hosts.eqiv and /etc/cmcluster has cluster ascii file on both the server.

Rajeev

[Karthik S S]

Try,

node1: vgchange -a n lockvg
node2: vgchange -a n lockvg
node1: vgchange -c n lockvg
node1: vgchange -a y lockvg
node1: cmapplyconf -C cluster.asc

Note that if you have activated the VG on both the nodes u may encounter problem during cmapplyconf.

Hopefully it should work now.

Regards,
Karthik S S

[Karamanolis Nikos]

Hi,

you should activate the cluster Lock vg, with
vgchange -a y lockvg

not the way you try it (i.e. -c y)

Hope it helps,

[Francesco Campalastri]

# cat test.ascii
CLUSTER_NAME clusterica
FIRST_CLUSTER_LOCK_VG /dev/vg01
NODE_NAME ica1
NETWORK_INTERFACE lan2
HEARTBEAT_IP 172.16.4.20
NETWORK_INTERFACE lan0
HEARTBEAT_IP 192.0.0.1
NETWORK_INTERFACE lan1
FIRST_CLUSTER_LOCK_PV /dev/dsk/c12t0d0
NODE_NAME ica2
NETWORK_INTERFACE lan2
HEARTBEAT_IP 172.16.4.40
NETWORK_INTERFACE lan0
HEARTBEAT_IP 192.0.0.2
NETWORK_INTERFACE lan1
FIRST_CLUSTER_LOCK_PV /dev/dsk/c21t0d0
HEARTBEAT_INTERVAL 1000000
NODE_TIMEOUT 2000000
AUTO_START_TIMEOUT 600000000
NETWORK_POLLING_INTERVAL 2000000
MAX_CONFIGURED_PACKAGES 10

# vgchange -a y vg01
Activated volume group
Volume group "vg01" has been successfully changed.
# cmapplyconf -C /etc/cmcluster/test.ascii

Begin cluster verification...
Note : a NODE_TIMEOUT value of 2000000 was found in line 18. For a
significant portion of installations, a higher setting is more appropriate.
Refer to the comments in the cluster configuration ascii file or ServiceGuard
manual for more information on this parameter.
Error: First cluster lock volume group /dev/vg01 needs to be designated as a clu
ster aware volume group.
cmapplyconf : Unable to verify cluster file: /etc/cmcluster/test.ascii.
#



Thanks again to everyone.
New ideas?

[melvyn burnard · Accepted solution]

try adding the line at the bottom of the ascii file:
VOLUME_GROUP /dev/vg01

You do not seem to have this entry, and for a cluster lock vg it must be defined in the ascii file

[Francesco Campalastri]

SOLVED!


Thanks to Melvyn

[melvyn burnard]

I LOVE it when a plan comes together :-]
Glad it worked.
Hope you also got the BEA stuff you asked for.