[Support Tip] In a Cluster local state shows offline even though the two cluster nodes show online

[GTrejos7]

PROBLEM:
[root@OMLSD1SAC bin]# ovclusterinfo -a
#Cluster
type         Red Hat Advanced Server (RHAS)
name         abc.net
status       Up
nodes        abc1.net abc2.net
groups       ov-server 
#Node abc1.net
state        Online
id           0 
#Node abc2.net
state        Online
id           1
#Group ov-server
state        Online
nodes        abc1.net abc2.net
local state  Offline
virtual IP   10.x.x.x
active node  abc1.net

As the local state output is checked during the installation of SPIs, this defect prevents proper installation of any SPI.

CAUSE:

The "local state  Offline" is wrong, it should be "Online".

SOLUTION:

The variable CLUSTER_LOCAL_NODENAME should be set on conf.cluster  (set this variable to the name which appears cluster config (probably not an FQDN...) ).

This variable should be set on both physical nodes:

abc1#ovconfchg -ns conf.cluster -set CLUSTER_LOCAL_NODENAME abc1.net

also on the second physical node:

abc2#ovconfchg -ns conf.cluster -set CLUSTER_LOCAL_NODENAME abc2.net

After this run again the command below on the active node to verify this HARG is now seen online

abc1.net#/opt/OV/bin/ovclusterinfo -group ov-server -localState"

There is already an ER to fix this automatically in future releases:

http://support.openview.hp.com/selfsolve/document/LID/QCCR1A104981