I am running RHEL5.5 and have a 2 node cluster that is setup with a quorum disk. In /var/log/messages I kept getting this message.

Sep 2 07:56:32 openmq2-dev qdiskd[4369]: <warning> qdisk cycle took more than 1 second to complete (1.210000)
Sep 2 07:56:41 openmq2-dev qdiskd[4369]: <warning> qdisk cycle took more than 1 second to complete (1.650000)
Sep 2 07:56:55 openmq2-dev qdiskd[4369]: <warning> qdisk cycle took more than 1 second to complete (1.960000)
Sep 2 07:57:13 openmq2-dev qdiskd[4369]: <warning> qdisk cycle took more than 1 second to complete (1.870000)
Sep 2 07:57:16 openmq2-dev qdiskd[4369]: <warning> qdisk cycle took more than 1 second to complete (1.420000)
Sep 2 07:57:30 openmq2-dev qdiskd[4369]: <warning> qdisk cycle took more than 1 second to complete (2.720000)
Sep 2 07:57:36 openmq2-dev qdiskd[4369]: <warning> qdisk cycle took more than 1 second to complete (2.660000)
Sep 2 07:57:44 openmq2-dev qdiskd[4369]: <warning> qdisk cycle took more than 1 second to complete (3.070000)
Sep 2 07:57:51 openmq2-dev qdiskd[4369]: <warning> qdisk cycle took more than 1 second to complete (1.900000)

I have looked on google but nothing that I finds tells me how to fix it. I think my cluster.conf is correct because it doesn't complain about anything. Any help is appreciated.

<?xml version="1.0"?>
<cluster alias="GFStestCluster" config_version="9" name="GFStestCluster">
<cman expected_votes="3" two_node="0"/>
<quorumd device="/dev/sdb" interval="1" min_score="1" tko="10" votes="1">
<heuristic interval="2" program="ping -c3 -t2 10.6.191.191" score="1"/>
</quorumd>
<clusternodes>
<clusternode name="openmq1-dev" nodeid="1" votes="1">
<fence>
<method name="single">
<device name="openmq1-dev_ipmi"/>
</method>
</fence>
</clusternode>
<clusternode name="openmq2-dev" nodeid="2" votes="1">
<fence>
<method name="single">
<device name="openmq2-dev_ipmi"/>
</method>
</fence>
</clusternode>
</clusternodes>
<fencedevices>
<fencedevice agent="fence_ipmilan" ipaddr="10.6.191.191" login="nimdadod" name="openmq2-dev_ipmi" passwd="nuk0sama"/>
<fencedevice agent="fence_ipmilan" ipaddr="10.6.191.192" login="nimdadod" name="openmq1-dev_ipmi" passwd="nuk0sama"/>
</fencedevices>
<rm>
<failoverdomains/>
<resources/>
</rm>
<fence_daemon post_fail_delay="0" post_join_delay="3"/>
</cluster>

Increase your quorumd interval to 7.

That's the number used in this Red Hat document. Though I admit I don't know how long is typical for the quorum operation to take place and one of your was over 3 seconds. If everything seems fine with your disk I would just increase the interval to 7 and not worry about it.

http://docs.redhat.com/docs/en-US/Re...iguration.html