This article has been archived. It is offered "as is" and will no longer be updated.
When you fail over an Exchange 2000 virtual server to another node on an Active/Passive cluster, and then fail back to the original node, you experience more virtual memory fragmentation on the original cluster node than you might expect.
This behavior occurs because the Exchange 2000-related services are not stopped when you fail over an Exchange 2000 virtual server to another cluster node. When an Exchange virtual server is failed over from one cluster node to another, the Exchange 2000-related services are started on the second cluster node if they are not already running. However, the Exchange 2000-related services are not automatically stopped on the node from which the Exchange 2000 virtual server failed over. Because of this, when the Exchange 2000 virtual server is failed back to the original node, virtual memory fragmentation of the information store (Store.exe) process may occur more rapidly than you might expect.
To work around this problem, before you move an Exchange 2000 virtual server back to a passive cluster node that already has the Exchange 2000 services running, restart the Exchange 2000-related services or that cluster node.
Microsoft has confirmed that this is a problem in the Microsoft products that are listed at the beginning of this article.
For additional information, click the following article number to view the article in the Microsoft Knowledge Base:
296073 XADM: Monitoring for Exchange 2000 Memory Fragmentation