Shutting down manager due to excessive heap growth

Anand informed us today about an error message in concurrent manager logs:

w1234567.mgr:Shutting down manager due to excessive heap growth:
w1234568.mgr: Orginal Top of Heap: 5708512
w1234679.mgr: Current Top of Heap: 35748576

Excessive heap growth is a common reason for a Java application performance problem.
Heap growth may be caused by your application having a memory leak or of the garbage
collection not doing its job. Memory leaks in the Java language are a significant contributor to
garbage collection bottlenecks. If you do not manage garbage collection, it can have a
significant negative impact on application performance, especially when running on
symmetric multiprocessing (SMP) server machines. The Java Virtual Machine (JVM)
uses concurrent (asynchronous) garbage collection. This type of garbage collection results in
shorter pause times and enables application threads to continue processing requests during
the garbage collection cycle.
Some symptoms of heap growth include:

• Poorly performing application that may have a leak
• Heavy page faulting in the storage pool where your WebSphere application runs (Java does not tolerate heavy paging)
  

You can use several tools to evaluate a possible heap growth problem:

• The Dump Java Virtual Machine (DMPJVM) command
• Start Service Tools (SST)
• Work with System Status (WRKSYSSTS) to investigate faulting and the wait-to-ineligible transitions.
• IBM's iDoctor for iSeries Heap Analyzer

This could be another post ATG RUP6 bug.