TIBCO ActiveMatrix BusinessWorks Engine Tuning Guidelines

This section provides high-level steps for TIBCO ActiveMatrix BusinessWorks™ performance tuning. Specific details are discussed in the subsequent sections.

When tuning the ActiveMatrix BusinessWorks™ engine allocate significant time to select the JVM configurations and the necessary tuning parameters depending on the use cases, payload and workload requirements.
  • Since jobs process on a separate Java thread, the number of engine threads controls how many jobs can run simultaneously. The number of threads that an engine will allocate is set in the AppNode config .ini file
  • Measuring the available CPU and memory resources on your system under a typical processing load will help you to determine if the default value of eight threads is appropriate for your environment.

    For example, if engine throughput has reached a plateau, yet measurements show that CPU and memory are not fully utilized, increasing this value can improve throughput. Typically it is advisable to double the engine threads when tuning the engine if CPU resources are available.

  • There are three ways of using CPU resources to the maximum capacity:
    • Optimize engine thread.
    • Increase number of AppNodes. Your performance suite must cover different scenarios so that you can come up with the optimal configuration.
    • Tune thread pools, where available, for asynchronous activities.
  • Typical numbers of engine threads range between eight and thirty-two. Specifying a value too low can cause lower engine throughput even though spare CPU resources exist. Specifying a value too high can cause CPU thrashing behavior.
  • If the pace of the incoming processes still exceeds the number of threads available to run them, consider using flow control.