Microsoft released a hotfix for failover clustering on Windows Server 2008 and Windows Server 2008 R2.
Assume that you set up a failover cluster in a Windows Server 2008 or Windows Server 2008 R2 environment. However, after the Cluster service runs for a long time (for example, 100 days), the service stops. Additionally, the following error is logged in the Cluster event log:
Workitem callback threw exception: InternalError(1359)’ because of ‘internal error'(Reply to can only be invoked for messages sent by Mrr component).
This issue occurs because the value of the Multicast-Request-Reply (MRR) ID changes to -1. This value cannot be processed.
A supported hotfix is available from Microsoft.