Cluster Resources May Go Offline After Event ID 3013 (308151)

MORE INFORMATION

Windows Clustering uses a mechanism called "LooksAlive" to periodically check to determine if a clustered resource is still online. For a file share, the Cluster service traverses the network path to the file share to determine if the resource is still online. However, if event ID 3013 occurs (which is a redirector time-out), the network path is no longer available because of the redirector time-outs. Eventually, when the redirector is available, the file-share resources come back online because the LooksAlive test is then successful.

By default, a LooksAlive test occurs every 5,000 milliseconds. You can increase this value by using the Cluster Administrator tool, but doing so only treats the symptoms, not the cause of the redirector time-outs.

Event ID 3013 may also occur with event ID 2022 ("Server was unable to find a free connection X times in the last 60 seconds"). This occurs if the Server service is unable to keep up with the system demand for connections; this can be caused by disk stress because of low disk space, extreme fragmentation, or both, coupled with high disk usage. Under these conditions, it is likely that both 3013 and 2022 events will occur.

The fact that cluster resources go offline is a symptom of the root problem that is causing the 3013 and 2022 events. To ensure the stability of the cluster, you must resolve the root problem that is causing the events by methods such as adding disk resources, installing faster controllers, and so on. For additional information about resolving these events, click the article numbers below to view the articles in the Microsoft Knowledge Base: