Workflow for performance tuning

You should follow the workflow outlined in this topic whenever you encounter a performance problem.

1. Determine if you have a performance problem. A performance problem is defined as a situation where the latency you are experiencing in your production environment does not meet your business requirements.
   Important: You cannot use your test environment to identify performance problems unless it is identical to your production environment. For example, the hardware, software (database and other applications), database workload, and all other aspects of the environment must be identical.
2. Analyze all of the tables and columns that you are replicating and determine if there is business value in these tables and columns. You should remove all tables and columns with no business value. For more information, see Assessing business value for tables, columns, and rows.
3. Assess the health of they system where IBM® Data Replication is deployed. For more information, see Performance tuning your system for CDC Replication.
4. Use the Management Console Performance view to identify the bottleneck that is causing the latency. For more information on how to use this tool, see Using Management Console to identify performance bottlenecks.
5. Once you have identified the bottleneck that is causing your performance problem, go to the section of this guide that is devoted to the type of bottleneck you are experiencing. These sections contain information on the type of configurations and functionality that may cause an undesirable level of latency. You may have to make changes to your environment or your CDC Replication configuration to reduce your levels of latency so that it meets your business requirements.
6. After resolving your bottleneck, you may still have a performance problem because of a different type of bottleneck. If this is the case, return to Step 1 of this procedure.