How to: Verify Team Foundation Server Fail-Over in a Cluster

  • Article

By inspecting the status of the Team Foundation services on the application tier, you can verify that the cluster and the generic script resource work if the active node is taken offline. The procedure makes the following assumptions: Team Foundation Server is deployed in a complex topology that has the data tier in a cluster and not a single- or dual-server topology. The cluster administrator has added a generic script resource that resets the services on the application tier. If you do not have the generic script resource added to the cluster, you might experience a significant delay in recovery of services.

Required Permissions

To perform this procedure, you must be a member of the Cluster Administrator security group. For more information, see Team Foundation Server Permissions.

In addition to these permissions, you might need to address the following requirements on a computer that is running Windows Server 2008 or Windows Vista:

  • To follow a command-line procedure, you might need to open an elevated Command Prompt by clicking Start, right-clicking Command Prompt, and clicking Run as Administrator.

  • To follow a procedure that requires Internet Explorer, you might need to start it as an administrator by clicking Start, clicking All Programs, right-clicking Internet Explorer, and then clicking Run as administrator.

  • To edit web.config files, you might need to start the text editor as an administrator by clicking Start, clicking All Programs, right-clicking the editor, and then clicking Run as administrator.

  • To access Report Manager, reports, or Web sites for SQL Server Reporting Services, you might need to add these sites to the list of trusted sites in Internet Explorer or start Internet Explorer as an administrator.

For more information, see the Microsoft Web site.

To take a node offline and test recovery

  1. On the application tier, open Windows Task Manager by right-clicking an empty area on the taskbar, and then click Task Manager.

  2. On the Processes tab, verify that IIS is running by locating w3wp in the Image Name column.

  3. On the active node of the data-tier cluster, open Cluster Administrator by clicking Start and then clicking Cluster Administrator.

  4. On the File menu, click Initiate Failover.

  5. On the application tier, open Windows Task Manager and verify that IIS w3wp no longer appears on the Processes tab.

  6. In the Cluster Administrator on the data tier, right-click the script resource and select bring online.

    Note

    The cluster may not have a script resource to explicitly restart the services. If not, you can add one by using the procedure, "How to: Accelerate Service Recovery in a Cluster" in the Team Foundation Server Installation Guide. Without this script, you may experience a significant delay before recovery has completed.

  7. In Internet Explorer on the application tier, locate the Web service and verify that GetRegistrationEntries is available.

    For example, you can enter the following Internet address to view the registration information for the application tier installed on the server ADATUM1.

    http://ADATUM1:8080/services/v1.0/registration.asmx

Note

For more information about how to test failover, see whether group resources can fail over at "Single Quorum Device Server Cluster Configuration" (https://go.microsoft.com/fwlink/?LinkId=62449).

See Also

Concepts

Ensuring Team Foundation Server Availability

Clustering the Data-Tier Server

Other Resources

Managing Data