The Service Integration Bus Performance tool is a stand-alone program specifically designed to help with the understanding of messaging performance. It automatically preselects and displays key messaging statistics.

Running the tool with default settings would provide the following information (automatically updated every two seconds):

• current message production and consumption rate for every queue and topic on the system (including temporary queues)
• message production and consumption rates broken down by reliability level for each queue and topic
• current number of producers/publishers and consumers/subscribers attached to each queue/topic
• number of messages, using the Best Effort reliability level, that have been discarded
• numbers of threads in each thread pool (for example: MDBs, ORB, Web container)
• number of bytes written and received by the application servers, message engines, and MQ links
• heap size and current percentage of free heap.

How does it work?

The tool is written entirely in Java™ using the PMI (Performance Monitoring Interface) APIs provided by the WebSphere® Application Server. The look and feel of the GUI is provided by the Standard Widget Toolkit, which was developed by the Eclipse foundation.

David Granshaw is a senior engineer, performance analyst, and team lead for the Service Integration Bus Performance team. He had ten years of experience as a scientist before joining IBM in 1998. At IBM, he has worked in Java service, ORB development, ORB performance, Java performance, and Service Integration Bus Performance.