Trying to configure 2-node cluster on HP Porliant BL685c servers with RHEL 4.5. Once service ccsd is started getting error message in /var/log/messages saying,"unable to connect to infrastucture after 30 seconds.". This is happening every 30 seconds. Checked everything but no luck so far to get rid of this error and continue. These servers are not under any firewall setup. Please help me to troubleshoot this issue.

-regards
Raj

Normally this only happens if your cluster isn't quorate. Did you start cman? And what does your cluster.conf look like?

cman is starting but fenced is not. My cluster.conf file is as follows:
<?xml version="1.0" ?>
<cluster alias="hsestg_cluster" config_version="3" name="hsestg_cluster">
<fence_daemon post_fail_delay="0" post_join_delay="600"/>
<clusternodes>
<clusternode name="ma5orcl701aa.corp.halliburton.com" votes="1">
<fence>
<method name="1">
<device name="ma5orcl701_ILO"/>
</method>
</fence>
</clusternode>
<clusternode name="ma5orcl702aa.corp.halliburton.com" votes="1">
<fence>
<method name="1">
<device name="ma5orcl702_ILO"/>
</method>
</fence>
</clusternode>
</clusternodes>
<cman expected_votes="1" two_node="1" log_facility="local7" log_level="7"/>
<ccsd log_facility="local7" log_level="7"/>
<fencedevices>
<fencedevice agent="fence_ilo" hostname="ma5orcl701_ilo.corp.halliburton.com" login="Administrator" name="ma5orcl701_ILO" passwd="init2007"/>
<fencedevice agent="fence_ilo" hostname="ma5orcl702_ilo.corp.halliburton.com" login="Administrator" name="ma5orcl702_ILO" passwd="init2007"/>
</fencedevices>
<rm>
<failoverdomains/>
<resources/>
</rm>
</cluster>

When starting the fenced daemon the following error message is displayed:
Jul 9 01:33:13 ma5orcl701aa ccsd[30523]: Unable to connect to cluster infrastructure after 180 seconds.
Jul 9 01:33:34 ma5orcl701aa ccsd[30523]: Cluster is not quorate. Refusing connection.
Jul 9 01:33:34 ma5orcl701aa ccsd[30523]: Error while processing connect: Connection refused
Jul 9 01:33:35 ma5orcl701aa ccsd[30523]: Cluster is not quorate. Refusing connection.
Jul 9 01:33:35 ma5orcl701aa ccsd[30523]: Error while processing connect: Connection refused

Please let me know if you could think of anything.

Hmm, everything looks quite ok.
I would like to see more messages from cman. And what does fenced tell you when you start it (fence_tool join).
Did you already tell about the versions and distribution you are using?
What does cman_tool status and cman_tool nodes tell you?
Also try to check if the hostnames resolve to ips. Best via /etc/hosts.

That's all for now.
Good look and don't loose patience
MG.

Another thing to not forget:

For a two node cluster both nodes need to be up and joined to the cluster before it gets quorate.

Sorry for the delay. I was busy with some other projects and would be starting on this one either today or tomorrow and would post more info.
-Raj

I sent ages for this issue. Finally, I got the point. It is because yum command is hung. You can check by

ps aux | grep yum

You will see at least 1 yum process. Kill it/them by

killall yum

Good luck