Cluster Migrations Involving New Storage: Mount Points

Updated: January 7, 2010

Applies To: Windows Server 2008 R2

This topic describes considerations for mount points in a migration to a failover cluster running Windows Server 2008 R2 where you use new storage for the migrated cluster. This is in contrast with a scenario where you use the same storage before and after the migration.

The Migrate a Cluster Wizard does not migrate mount point information (that is, information about hard disk drives that do not use drive letters, but are mounted instead in a folder on another hard disk drive). However, it can migrate Physical Disk Resource settings to and from disks that use mount points. The wizard also does not configure the necessary dependency between the resources for mounted disks and the resource for a host disk (the disk on which the other disks are mounted). You must configure the dependency after the wizard finishes running.

When you are working with new storage for your cluster migration, you have some flexibility in the order in which you complete the tasks. The tasks that you must complete include creating the mount points, running the Migrate a Cluster Wizard, copying the data to the new storage, and confirming the disk letters and mount points for the new storage. After completing the other tasks, configure the disk resource dependencies in Failover Cluster Manager.

A useful way to keep track of disks in the new storage is to give them labels that indicate your intended mount point configuration. For example, in the new storage, when you are mounting a new disk in a folder called \Mount1-1 on another disk, you can also label the mounted disk as Mount1-1. (This assumes that the label Mount1-1 is not already in use in the old storage.) Then when you run the Migrate a Cluster Wizard and you need to specify that disk for a particular migrated resource, you can look at the list and select the disk labeled Mount1-1. Then you can return to Failover Cluster Manager to configure the disk resource for Mount1-1 so that it is dependent on the appropriate resource, for example, the resource for disk F. Similarly, you would configure the disk resources for all other disks mounted on disk F so that they depended on the disk resource for disk F. The following illustrations show selected stages in this process.

In the first illustration, a new disk has been labeled Mount1-1 (through Disk Management), and has been added to the cluster Available Storage. The new disk is shown in Failover Cluster Manager.

Disk in Available Storage labeled Mount1-1

For the next illustration, the Migrate a Cluster Wizard has been used to specify that new, labeled disks in Available Storage are to be used in the new configuration, instead of the old disks being used. After this has been specified, the Migrate a Cluster Wizard presents a display similar to the following:

Migration with mounted disks in new storage

After you run the wizard and fully configure the mounted disk, the last step you need to take is to configure the disk dependencies in Failover Cluster Manager. For each disk resource for a mounted hard disk drive, open the Properties sheet, and on the Dependencies tab, specify a dependency on the disk resource for the host drive (where the mounted drives reside). This ensures that the Cluster service brings the host drive online first, followed by the drives that are dependent on it.

After you configure the dependencies, you can view a dependency report as follows: Click the service or application, and then under Actions, click Show Dependency Report. The following illustration shows four mount points that are configured with the correct dependencies on the disk on which they are mounted:

Four mount points with dependencies configured

Additional references

Migrating Clustered Services and Applications to Windows Server 2008 R2 Step-by-Step Guide