Share via


Configure Message Queuing as a Highly Available Cluster Resource

Applies To: Windows Server 2008

Use this procedure to configure Message Queuing as a highly available cluster resource.

You can use this procedure to configure Message Queuing as a highly available cluster resource.

Membership in the local Administrators group, or equivalent, is the minimum required to complete this procedure.

To configure Message Queuing as a highly available cluster resource

  1. Click Start, point to AllPrograms (or Programs if using Classic Start menu), point to Administrative Tools, and then click Failover Clusters Manager to display the Failover Cluster Management console.

Note

If you are not logged on a member of the local Administrators group then you must right-click Failover Clusters Manager and click Run as administrator or the Failover Clusters Manager will not start.

  1. Select Manage a Cluster from the Action menu.

  2. Select the applicable cluster name from the Cluster name dropdown box and click OK.

  3. Click to expand the cluster group, right click Services and Applications, and then click Configure a Service or Application to start the High Availability Wizard.

  4. Select Message Queuing from the list of services and applications and click Next.

  5. Enter a name into the Name edit box. This will be the NetBIOS name created in Active Directory Domain Services that is used by the network name resource which is created for the clustered Message Queuing service.

  6. The High Availability Wizard will attempt to automatically configure IP address resources when clustering Message Queuing. If automatic configuration does not succeed then you will be prompted to enter a valid IP address. If prompted, enter an IP address under Address and click Next. This will be the IP address used by the IP Address resource that is created for the clustered Message Queuing service.

  7. Select a storage volume to be used by the clustered Message Queuing service and click Next.

Note

The Select Storage dialog box of the High Availability Wizard lists all clustered disk resources that are available for use on the cluster. If there are not any clustered disk resources listed then you must cancel out of the High Availability Wizard, create a clustered disk resource, and start the High Availability Wizard again.

  1. Click Next on the Confirmation page of the High Availability Wizard.

  2. The High Availability Wizard will attempt to configure Message Queuing as a highly available resource using the resources that were provided. The High Availability Wizard will display the status of this attempt on the Completion page. If the attempt failed for any reason, click View Report for additional details on why the attempt failed and take corrective actions.

  3. Click Finish to close the High Availability Wizard.

  4. Reboot the cluster node.

Additional considerations

  • This task is performed on a computer that is running as part of a server cluster, and has the Message Queuing service already installed.

  • Errors occurring during the performance of this task are recorded in a Windows event log. For example, a "device not ready" error indicates that Message Queuing was not installed on the node.

  • All Message Queuing applications that run in the context of a cluster group (whether cluster-aware or not) must depend on the Message Queuing resource in that group, and use the Network Name resource in that group. Refer to the Server clustering documentation for information about creating Generic Application resources for "cluster-unaware" applications.

  • When you select the Physical Disk resource in the group, Message Queuing allocates its storage in the MSMQ\STORAGE folder on the shared disk. After storage has been allocated, you cannot modify the folder location.

  • To manage Message Queuing in the applicable resource group, right-click the resource group and select Manage MSMQ on the hosting node.

  • To uninstall Message Queuing on a computer that is part of a server cluster, the Message Queuing virtual server must be deleted from the node before uninstalling Message Queuing on the physical node. For example, if Message Queuing is installed on NodeA of a server cluster, you need to identify all Message Queuing resources on this cluster, ensure that NodeA does not own any cluster groups containing Message Queuing resources, and remove NodeA from the list of possible owners of each Message Queuing resource. Then you uninstall Message Queuing from NodeA.

Additional references