Creating and scheduling query service administration tasks

Administrators can create and schedule query service tasks for data sources. Query service tasks control one or more cubes by clearing, writing, or refreshing its cache. For dynamic cubes, you can also schedule when cubes start, stop, or restart, and refresh security.

schedule cache clearing and clear the cache to control memory usage by a specific data source or cube
schedule the generation of a time-stamped report (write cache state)

You can also clear the entire cache manually and write the cache state to a report manually.

For more information, see Clear everything in the cache and Analyzing cache usage.

You can create query service administration tasks and run them on demand. You can run them at a scheduled time Schedule Management or based on a trigger, such as a database refresh or an email Trigger-based Entry Scheduling. You can schedule them as part of a job Use Jobs to Schedule Multiple Entries. You can also view the run history of query service administration tasks View the Run History for Entries.

Before you begin

When you create and schedule tasks for dynamic cubes, you need to schedule start and stop tasks for source cubes and virtual cubes separately. There are other factors to consider when scheduling start and stop tasks for dynamic cubes:

Source cubes that are a part of a virtual cube must be scheduled to start first.
If source cubes are part of a virtual cube, then the virtual cube must be scheduled to stop before the source cubes.
You need to provide enough time for source cubes to start before scheduling a virtual cube to start. The same consideration must be made when you schedule virtual and source cubes to stop.

Procedure

In IBM Cognos Administration, on the Configuration tab, click Content Administration.
Click the New Query service administration task button .
Specify a name, description, screen tip, and location. Click Next.
Select an operation, either Clear Cache or Write Cache State.
For Oracle Essbase and SAP BW data sources, enter the data source, catalog, and cube. Click Next.
Enter an asterisk (*) as a wildcard to specify all.
For Dimensionally-Modeled Relational (DMR) data sources, enter either the name of a package name or the name of a data source. If you specify a data source name and chose the Clear Cache operation, the cache is cleared for all packages that involve that data source.
For dynamic cube tasks, select the Server Group, Dispatcher, and Cubes, and then click Next.
Choose the action that you want:
- To run the task now or later, click Save and run once and click Finish. Specify a time and date for the run, and then click Run. Review the run time and click OK.
- To schedule the task at a recurring time, click Save and schedule and click Finish. Then, select frequency and start and end dates. Click OK.
  Tip: To temporarily disable the schedule, select the Disable the schedule check box.
  
  Note: To view the schedule status, see Manage Scheduled Activities.
- To save the task without scheduling or running, click Save only and click Finish.

What to do next

You must remember to delete a scheduled task if you delete the associated cube from the query service. Otherwise, your scheduled tasks will point to nonexistent cubes.