Running Siebel EIM Tasks in Parallel

This topic is part of Troubleshooting Siebel EIM Performance.

Running Siebel EIM tasks in parallel is the last strategy that you might want to adopt in order to increase the Siebel EIM throughput rate. In other words, do not try this until all long-running SQL statements have been tuned, the optimal batch size has been determined, the optimal number of records to be processed at a time in the EIM table has been determined, and the database has been appropriately tuned. Before running tasks in parallel, check the value of the Maximum Tasks parameter. This parameter specifies the maximum number of running tasks that can be run at a time for a service. For more information about this parameter, see Siebel System Administration Guide.

Note: UPDATE STATISTICS must be set to FALSE in the IFB file when running parallel Siebel EIM tasks on IBM DB2 for z/OS. Failure to do so can cause Siebel EIM tasks and executing RUNSTATS to take a longer time to complete. Also, when running parallel Siebel EIM tasks, deadlock and timeout will occur if UPDATE STATISTICS is set to TRUE in the IFB file.