Operational Planning
Bulk Plan Performance Tuning
There are several reasons why some bulk plans run slowly, while other bulk plans, even with larger data sets run fast. For example, high run time may be warranted due to the nature of the underlying problem and the complexity of the network. Many performance problems though can be resolved through careful data modeling, algorithm tuning and multi-thread property settings.
One of the most important contributors to long run time is the amount of logs that are turned on. It is recommended that logs be used during the test phase, and turned off during production runs. Remember to turn off logs from all the domains, as the logs from other domains will influence the run time even though no logs are printed.
- Understanding the Bulk Plan Run provides an overview of the bulk plan process as well as key areas to check for performance.
- Bulk Plan Multi-threading is one way to take advantage of multiple CPUs when running bulk plans. You can run several bulk plans simultaneously by setting up multiple batch threads or by bulk plan partitioning.
- Multi-threading Logic explains how multi-threading works.
- Bulk Plan Partitioning is a way to enhance the overall performance of the bulk plan run.
- Bulk Plan Caches help reduce database access and computation times significantly and improve bulk plan performance.