Month: January 2013

Microsoft issued another clustering update for Windows Server 2008 R2.  This one is where:

The Cluster service still uses the default regroup time-out value of 6 seconds after you configure the regroup time-out setting on a computer that is running Windows Server 2008 R2. This time-out value may not be long enough for some cluster nodes to rejoin a large scale cluster after you restart the Cluster service.
Note To obtain the time-out value (ClusSvcRegroupOpeningTimeout), type the following command at a command prompt, and then press Enter:

cluster /prop

Microsoft says you can fix this by Installing the available hotfix.

After you install the following hotfix package, the following regroup settings become effective:

• ClusSvcRegroupOpeningTimeout
• ClusSvcRegroupPruningTimeout
• ClusSvcRegroupStageTimeout
• ClusSvcRegroupTickInMilliseconds

Microsoft issued a hotfix over the holidays to deal with an issue where:

• You configure a failover cluster that has some computers that are running Windows Server 2008 R2.
• One cluster node is offline or is removed from the cluster.
• You try to rejoin the cluster node to the cluster after a while.

In this scenario, the cluster node cannot rejoin the cluster.

According to Microsoft, the cause is that:

When a cluster node tries to rejoin a cluster after it is restarted or removed from the cluster, the virtual IP address of the cluster node may be changed. However, other cluster nodes still use old route records in their route database. Therefore, the cluster node cannot rejoin the cluster.

A supported hotfix is available from Microsoft.