Cluster Service Group Does Not Fail Over

Latest response

Hi all,

I have a two node RedHat Cluster with one Service Group configured.
The servers are running RHEL 6.2 on two VMware virtual machines.

My issue is that no matter how many times a resource fails, the Service Group never fails over to the other node, it just restarts on the same node, even though I tried configured the Failover Domain policy to restart (with a maximum of 1 restart) and relocate.

Whenever a resource fails, the logs show the cluster stop the group and starts an inline recovery (rgmanager[2788]: Some independent resources in service:App-SG failed; Attempting inline recovery).

I'm attaching my cluster.conf file.
I've removed the actual IP address from the file obviously.

I'd appreciate it if someone could help me understand what's incorrect in my configuration.
Thanks.

Attachments

YZ
Log in to join the conversation

Responses

rhn-support-mflitter's picture Red Hat Guru 1638 points
17 July 2014 11:45 AM

Hello Yair,

I am not a cluster expert, but it looks very much like you are hitting a variant of Why did my clustered service take a long time to fail or recover when the resources have the "__independent_subtree" attribute enabled on RHEL 5 or RHEL 6?.

I think you need to look at example 2, please let us know if that helps,
br
Mark

YZ Community Member 75 points
17 July 2014 12:32 PM

Hi Mark,

Thanks for your help.
I looked at the examples and explanations in the link you provided, however, it only made me more confused regarding why I should use an independent subtree and how can I cause it to failover between servers.

In any case, I have removed the independent subtree configuration from the IP address, and the failover process is according to the what I've planned.
Thanks again for you help.

Yair

Close

Welcome! Check out the Getting Started with Red Hat page for quick tours and guides for common tasks.