I am new to SQL Server Clustering and just worked with a partner on configuring a two node cluster on a Windows 2003 server and SQL Server 2000 in an Active/Passive design with one instance. While performing the fail-over tests to check that the cluster is working properly we began to fail over the individual groups. It just so happened that the customer was near by to observe our fail-over test and raised the following concerns. During the Failing over of the separate groups, the system is in an Active/Active state for the moment because the groups are split between the nodes. Theoretically, I think the customer is correct but it shouldn#%92t be a concern because fail-over is not considered complete until all groups have failed over. I think the question is not relevant. Please provide feedback on this. Is there a way to create dependencies on the groups that when one fails over all groups will automatically follow? Although I think the customer is a bit paranoid, this sounds like a valid question. On the other hand, I think when you need to perform maintenance on one node, just fail-over the entire node. What do you think? Note: These question are based on the manual fail-over test during the configuration of the windows and sql server cluster. Hope this helps a bit.