Using JConsole to monitor workflow memory usage

Building workflows (using Taverna for example) can sometimes run into Java heap space problems. I have had problems in the past where the workflow would hang and the following error would be generated on the command line:

java.lang.OutOfMemoryError: Java heap space

Obviously if you are sending large jobs through a workflow with numerous processors this problem will occur. What I have found really useful is the java tool JConsole which allows you to monitor java heap space, memory usage, threads, etc.

I had a workflow that kept running into java heap space problems whereby the committed memory was exceeding 1000MB. I altered the run.me script in Taverna to allocate 2048M of memory for Java allowing the workflow to complete successfully.

Fig.1. JConsole monitoring Java heap space.