Redhat Cluster Implementation query
 

Hi,

I hope you all are doing fine.

I have some problem implementing redhat cluster.

I have implemented the redhat cluster and it is showing the running nodes. I am using luci interface to look upon the status of the node in cluster.

My cluster configuration is as follows:-

<?xml version="1.0"?>
<cluster config_version="12" name="DemoCluster">
<clusternodes>
<clusternode name="node1" nodeid="1">
<fence>
<method name="Node1_method"/>
</fence>
</clusternode>
<clusternode name="node2" nodeid="2"/>
</clusternodes>
<cman expected_votes="1" two_node="1"/>
<rm>
<failoverdomains>
<failoverdomain name="DemoFailover" nofailback="1" ordered="1">
<failoverdomainnode name="node1" priority="10"/>
<failoverdomainnode name="node2" priority="11"/>
</failoverdomain>
</failoverdomains>
<resources>
<ip address="192.168.2.166" sleeptime="10"/>
<apache config_file="conf/httpd.conf" name="123" server_root="/etc/httpd" shutdown_wait="0"/>
</resources>
<service autostart="0" name="DemoService" recovery="relocate">
<ip ref="192.168.2.166"/>
</service>
<service autostart="0" name="456" recovery="relocate">
<apache ref="123"/>
</service>
</rm>
</cluster>


The main problem is that whenever i am starting the services it is not reflecting in the node, i mean it is not running the services. The log message of my node is :-

Dec 16 17:27:39 node2 ricci[6264]: Executing '/usr/bin/virsh nodeinfo'
Dec 16 17:27:40 node2 ricci[6266]: Executing '/usr/libexec/ricci/ricci-worker -f /var/lib/ricci/queue/1509398375'
Dec 16 17:27:40 node2 ntpd[2581]: synchronized to 111.93.148.118, stratum 2
Dec 16 17:27:41 node2 ricci[6269]: Executing '/usr/libexec/ricci/ricci-worker -f /var/lib/ricci/queue/2017267267'
Dec 16 17:27:44 node2 ricci[6280]: Executing '/usr/bin/virsh nodeinfo'
Dec 16 17:27:44 node2 ricci[6283]: Executing '/usr/bin/virsh nodeinfo'
Dec 16 17:27:54 node2 ricci[6322]: Executing '/usr/bin/virsh nodeinfo'
Dec 16 17:27:54 node2 ricci[6324]: Executing '/usr/libexec/ricci/ricci-worker -f /var/lib/ricci/queue/1462165413'
Dec 16 17:28:03 node2 ricci[6351]: Executing '/usr/bin/virsh nodeinfo'
Dec 16 17:28:03 node2 ricci[6353]: Executing '/usr/libexec/ricci/ricci-worker -f /var/lib/ricci/queue/1909536095'
Dec 16 17:28:03 node2 modcluster: Starting service: 456 on node
Dec 16 17:28:12 node2 ricci[6381]: Executing '/usr/bin/virsh nodeinfo'
Dec 16 17:28:12 node2 ricci[6383]: Executing '/usr/libexec/ricci/ricci-worker -f /var/lib/ricci/queue/1887274114'
Dec 16 17:28:13 node2 ricci[6386]: Executing '/usr/libexec/ricci/ricci-worker -f /var/lib/ricci/queue/1430758334'
Dec 16 17:28:17 node2 ricci[6398]: Executing '/usr/bin/virsh nodeinfo'
Dec 16 17:28:31 node2 ricci[6461]: Executing '/usr/bin/virsh nodeinfo'
Dec 16 17:28:31 node2 ricci[6465]: Executing '/usr/libexec/ricci/ricci-worker -f /var/lib/ricci/queue/1073142511'
Dec 16 17:28:40 node2 ricci[6646]: Executing '/usr/bin/virsh nodeinfo'
Dec 16 17:28:40 node2 ricci[6648]: Executing '/usr/libexec/ricci/ricci-worker -f /var/lib/ricci/queue/413796341'
Dec 16 17:28:40 node2 modcluster: Starting service: 456 on node node1
Dec 16 17:28:49 node2 ricci[6668]: Executing '/usr/bin/virsh nodeinfo'
Dec 16 17:28:49 node2 ricci[6670]: Executing '/usr/libexec/ricci/ricci-worker -f /var/lib/ricci/queue/846327886'
Dec 16 17:28:50 node2 ricci[6687]: Executing '/usr/libexec/ricci/ricci-worker -f /var/lib/ricci/queue/745333525'
Dec 16 17:28:54 node2 ricci[6698]: Executing '/usr/bin/virsh nodeinfo'

Could anyone please tell me the solution for the above.

It looks like you are trying to cluster libvirt? I'm lead to believe you are trying to cluster a virtual machine.

I know people try this, but more or less isn't the general idea behind clusters. The general idea being having services, not hosts, up all of the time.

Yes i am doing cluster between different VMs. Is it not possible to it or is it?

Thanks for the reply.

Apparently this is possible???? It would depend on what your virtualization you are using among other things.

http://clusterlabs.org/wiki/Guest_Fencing claims it does.

Digging around a bit, it looks like migration is tricky. (no surprise) I spend enough time keeping clustered apps going I haven't bothered trying to set up a cluster that migrates a VM.