Cluster setup may not work when you add nodes

When you add nodes to a cluster, The Setup Wizard may stop working during the Analyze phase. When this occurs, the following error message is generated: This occurs only while you are adding nodes to an existing cluster.

This behavior is caused by an interaction between two technologies: the new Cluster Setup Wizard and the storage area network (SAN) configuration. The new Cluster Setup Wizard contains heuristics to verify that all cluster nodes have access to the same disks. The wizard does this by determining whether all cluster nodes recognize the disks with the same target ID (TID) and logical unit number (LUN). Typically, in a correctly configured cluster shared-disk array, this is the case. However, with a complex SAN configuration, this may not be true.

With a complex SAN environment, the same set of disks may be detected on different LUNs and TIDs by individual nodes, but this can still be a valid cluster configuration. Therefore, the disk-detection heuristics may incorrectly detect that the disk configuration is not valid for Setup.

To work around this issue, click Advanced when you are prompted for the nodes to add to the cluster, and then click Advanced (minimum) configuration. The Advanced option excludes the detection that is described earlier in this article from the Analyze process, and permits the cluster installation to continue. However, if there are disk-configuration issues, the cluster may experience issues with bringing the disks online or with disk failover.

One example of a complex storage solution might be a SAN with multiple switches between the nodes and the storage. Because the cluster nodes do not require the same path to any disk, nodes may enumerate the disks with different IDs. For example, node A may enumerate Disk1 as LUN 1 TID 1. Because node B can be connected to a separate switch and controller in the SAN and still enumerate the same disk, the disk might be enumerated, for example, as LUN 2 TID 3.

The Cluster Service Setup log (Clcfgsrv.log) contains information about how the nodes enumerate the disks. This is a sample of the information in the Clcfgsrv.log file: Note that another node in the cluster enumerates the same physical drive differently:The path of the Clcfgsrv.log file is %systemroot%\System32\Logfiles\Cluster\Clcfgsrv.log. You can use the Clcfgsrv.log file to determine and compare how the nodes enumerate the disks. You can also use Device Manager to determine this. To determine the TID and LUN for a disk in Device Manager, follow these steps:

1. Right-click My Computer, and then click Properties.
2. On the Hardware tab, click Device Manager.
3. Expand the Disks entry in Device Manager.
4. Right-click a disk, and then click Properties. The General tab contains an entry named Location. This entry shows the LUN and TID for the disk.

You can use these methods to determine if your cluster nodes are enumerating disks with different TIDs and LUNs. For more information about this topic, see the "Storage Configuration Options" topic in Help and Support Center.