Monitoring clustered applications such as SQL Server or Exchange Server in an active/passive environment is very similar to the monitoring example discussed in Section A.4.5, Using SetResDependency in an Active/Active Cluster: checking an ASCII log file on a shared cluster disk in an active/active cluster environment. As with monitoring shared logical disks, the Operator Console or Control Center TreeView should include all of the physical nodes that compose the cluster. You should place the physical nodes for the cluster in a unique server group.
As an example of monitoring a clustered application in an active/passive cluster environment, assume you want to monitor SQL Server installed as a clustered application. For this example, the virtual server name is VirtualSQLServer, the physical nodes are ROME and PARIS, and the cluster’s shared disk is defined as drive F:.
To monitor SQL Server in an active/passive configuration:
Install the AppManager agent and the AppManager for Microsoft SQL Server module on both the ROME node and the PARIS node.
If necessary, run the Discovery_NT and Discovery_SQL Knowledge Scripts.
In the Operator Console or Control Center, create a server group in the Master view called VirtualSQLServer and add the ROME and PARIS servers to the group. For more information about creating server groups, see the Control Center User Guide for AppManager.
Run the AMAdmin_SetResDependency Knowledge Script on the server group for the cluster, for example VirtualSQLServer. In the Properties dialog box:
Set Knowledge Script category to SQL.
Set Required available resources to be F:
Click OK to run the job. By default, the job runs once on each computer in the server group to establish the dependency. You should only need to run the Knowledge Script again if you make changes to the cluster membership or the cluster resource.
The values you set in the Knowledge Script Properties dialog box identify jobs that should run only when the specified resource is owned by the current node. Otherwise, the job is inactive. Although using a drive letter as a cluster resource dependency is common, as shown in Section A.5, Monitoring Applications in an Active/Passive Cluster, you can set the dependency to be a required active service when monitoring clustered applications.
When monitoring a virtual SQL Server, for example, you can set the dependency based on the mssqlserver$virtualservername service rather than the logical drive. Run the AMAdmin_SetResDependency Knowledge Script on the computer PARIS. Specify the SQL job category and the dependency as the MSSqlServer service. SQL jobs run on the computer PARIS when it is the active node that owns the MSSqlServer service.
When specifying a resource dependency, identify a single resource, whether it be a physical drive or a service name, within the cluster group where the monitored application is installed.
Once you have established the resource dependency, you can monitor the cluster using SQL Knowledge Scripts. Run the SQL Knowledge Scripts to monitor the virtual SQL Server resources on the server group you created for the cluster.