Hi All,
I need some help in setting up Linux cluster.
OS: CentOS 6.5 x86_64bit
modcluster-0.16.2-28.el6.x86_64
clusterlib-3.0.12.1-59.el6_5.2.x86_64
lvm2-cluster-2.02.100-8.el6.x86_64
3 node cluster.
If all the 3 nodes are started at once, cluster starts fine all 3 nodes are joined to the cluster. But if either one node fails and restart, the failed node will not join the cluster when booting, it give error as Unregistering quorum device.
Quote:
Aug 04 07:50:53 qdiskd Quorum Partition: /dev/block/8:0 Label: cluster-vm-01
Aug 04 07:50:53 qdiskd Quorum Daemon Initializing
Aug 04 07:50:53 qdiskd I/O Size: 512 Page Size: 4096
Aug 04 07:50:56 qdiskd Heuristic: 'ip link show bond0 | head -n1 | grep -q UP' UP
Aug 04 07:50:56 qdiskd Heuristic: 'ip addr show bond0 | head -n1 | grep -q UP' UP
Aug 04 07:50:59 qdiskd Node 1 is UP
Aug 04 07:50:59 qdiskd Node 2 is UP
Aug 04 07:51:01 qdiskd Node 1 is the master
Aug 04 07:51:13 qdiskd Initial score 2/2
Aug 04 07:51:13 qdiskd Initialization complete
Aug 04 07:51:13 qdiskd Score sufficient for master operation (2/2; required=1); upgrading
Aug 04 07:51:58 qdiskd Unregistering quorum device.
|
But if I manually start cman on the failed node, it will join the cluster.
Code:
<?xml version="1.0"?>
<cluster config_version="74" name="cluster-vm-01">
<cman expected_votes="2" two_node="0"/>
<quorumd interval="2" label="cluster-vm-01" min_score="1" tko="10" votes="2">
<heuristic interval="2" program="ip link show bond0 | head -n1 | grep -q UP" score="1" tko="3"/>
<heuristic interval="2" program="ip addr show bond0 | head -n1 | grep -q UP" score="1" tko="3"/>
</quorumd>
<fence_daemon post_fail_delay="0" post_join_delay="30"/>
<totem consensus="38000" token="36000"/>
<clusternodes>
<clusternode name="node1120" nodeid="1" votes="1">
<fence/>
</clusternode>
<clusternode name="node1121" nodeid="2" votes="1">
<fence/>
</clusternode>
<clusternode name="node1122" nodeid="3" votes="1">
<fence/>
</clusternode>
</clusternodes>
<fencedevices>
<fencedevice agent="fence_ipmilan" ipaddr="node1120.ipmi" login="cluster" name="ipmi_node1120" passwd="cluster"/>
<fencedevice agent="fence_ipmilan" ipaddr="node1121.ipmi" login="cluster" name="ipmi_node1121" passwd="cluster"/>
<fencedevice agent="fence_ipmilan" ipaddr="node1122.ipmi" login="cluster" name="ipmi_node1122" passwd="cluster"/>
</fencedevices>
<logging debug="on" logfile_priority="info" syslog_facility="local4" syslog_priority="info" to_logfile="yes" to_syslog="yes">
<logging_daemon logfile="/var/log/cluster/qdiskd.log" name="qdiskd"/>
<logging_daemon logfile="/var/log/cluster/fenced.log" name="fenced"/>
<logging_daemon logfile="/var/log/cluster/dlm_controld.log" name="dlm_controld"/>
<logging_daemon logfile="/var/log/cluster/gfs_controld.log" name="gfs_controld"/>
<logging_daemon logfile="/var/log/cluster/rgmanager.log" name="rgmanager"/>
<logging_daemon logfile="/var/log/cluster/corosync.log" name="corosync"/>
</logging>
</cluster>
