Entourage cannot send messages or receive messages after you fail over an active node to a passive node in an Exchange 2007 cluster

In a Microsoft Exchange Server 2007 environment, the Microsoft Entourage client cannot send or receive messages after you fail over active node to passive node. This problem occurs if the following conditions are true:

• You fail over a clustered mailbox server (CMS) from one active node to the first passive node. Then, you fail over the CMS to any other passive node.
• There are multiple clustered mailbox servers in an Exchange Server 2007 single copy cluster (SCC) environment.
• You fail over another Exchange Server virtual server from an active node to the first passive node.

When this problem occurs, the Entourage client receives a 170 error code. The PROPFIND command from the client fails with a 305 error code. Additionally, the GET command fails with a 302 error code in the Internet Information Services (IIS) log files.

The first time that you fail over from an active node to�a passive node, a new W3wp.exe process is created when an Entourage user connects to his or her mailbox. If�the CMS fails back to another node, the W3wp.exe process in the passive node�continues to run. Additionally, if�any other CMS fails over to the first passive node, the CMS will�try to use the same W3wp.exe process that has cached the server name and other information for the previous CMS.

To work around this issue, restart the IIS admin service in a passive node as soon as any node becomes a passive node. Then, a new W3wp.exe process is created when Entourage users try to connect to their mailboxes after an active node fails over to the passive node.

Microsoft has confirmed that this is a problem in the Microsoft products that are listed in the "Applies to" section.

A fix for this problem is scheduled to be included in Microsoft Exchange Server 2007 Service Pack 1 (SP1).