Linux 2 Node Cluster Lucci / Ricci Centos 6.3 Woes ....
 

Hi All,
I am very new to clusters and have setup a 2 node cluster with Luci and Ricci on Centos 6.3.

I have got the cluster to the stage that If I add an IP address as a resource, I can restart one of the nodes, and the virtual ip will transfer to the other node and stay online.

However as soon as I add a child resource to the IP, such as apache web server (httpd) it will listen fine on the virtual ip, however if I restart one of the nodes in the cluster, not only does the webpage not stay online but now the virtual ip also fails. Fencing is enabled in a Vmware environment and seems to be working.

So as soon as a Child Resource is added, everything is broken... As I remove it everything is fine again. Tried it with another Resource, and not a Child Resource and it still broke the Virtual IP, Resource..

Any ideas?

Also do you guys use Luci and Ricci for you HA Clusters? Or are you guys using Corosync and Pacemaker or Heartbeat for HA on your "services / daemons"

Thanks

Ok Solved it changed the "Shutdown" wait time on the apache aka "child" resource from 0 to 10 seconds, next step is to get more that one resource working also shared storage functioning....