Workload management sample application

Comprehensive workload management features have been integrated into your Db2® data server with Db2 workload management, giving you finer control over activities, resources and performance, and deeper insight into how your system is running. A workload management sample application is now available on developerWorks®.

The workload management sample application demonstrates how you can use Db2 workload management features to achieve the following objectives:

Protect the system from runaway queries
Runaway queries are costly and cause poor performance. The workload management sample application identifies queries with the potential to become runaway queries, and then stops these queries from running after they have violated a specified threshold.
Limit concurrent resource consumption by individual applications
The sample application shows how to use Db2 workload management features to prevent applications that submit large amounts of concurrent work from negatively affecting the performance of other applications.
Achieve a specific response time
Workload management features permit you to achieve a specific response time objective of the form: "transaction X from application Y shall complete within 1 second in 90% of cases," regardless of what other activity is running concurrently on the system. The sample application will demonstrate how to achieve a response time objective.
Consistent response time for short queries
Queries that typically have a response time of less than 1 second should have a relatively consistent response time regardless of what other workloads are running on the system. The sample application uses the query execution time histogram to monitor consistency.
Protect the system during periods of peak demand
Workload management policy features protect the system from capacity overload during bursts of peak demand by queuing work once the system is sufficiently loaded.
Enable concurrent batch extract, transform, and load (ETL) processing and user queries
Workload management features permit you to run ETL jobs (like loading data into tables) while controlling the performance impact for users running queries concurrently.

To obtain the sample application, see Workload management sample on developerWorks.