Skip to Content
avatar image
Former Member

Cluster failover

I have a couple of clusters producing unexpected behaviour. The problem is the same for both.

Two nodes in the cluster, all resources are on Node A. SAP section of the cluster fails over to Node B for no apparent reason, SQL stays on A.

Node A and Node B can see each other no problem. However, users cannot connect to the B Node, only A.

Node B doesn't failback and if you try to make it by using Move Group it fails and switches back to Node B.

To make it move back you have to shut SAP on Node B and Node A. It will then let you move it back.

Any help or advice much appreciated.

David

Add comment
10|10000 characters needed characters exceeded

  • Get RSS Feed

3 Answers

  • avatar image
    Former Member
    Aug 20, 2007 at 11:16 AM

    First response question: What do the windows logs say?

    Add comment
    10|10000 characters needed characters exceeded

  • avatar image
    Former Member
    Aug 20, 2007 at 11:20 AM

    that are two problems:

    1. unexpected failover

    2. move back to A did not work

    I need a little more information for getting the right answers.

    - SAP release

    - what did you find in the system log (event viewer, system)

    - assuming you are running on 2003 output of tasklist.exe on A before you move the resource back to A.

    peter

    Add comment
    10|10000 characters needed characters exceeded

  • avatar image
    Former Member
    Aug 20, 2007 at 12:06 PM

    Thank you the quick response.

    Following some greater investigation into the Event logs I think this may not be a SAP issue after all but rather a Windows/Cisco issue with the Shares timing out.

    David

    Add comment
    10|10000 characters needed characters exceeded