Setting "Max Active" on a memory pool.

I'm not quite sure I fully understand what "Max Active" on a memory pool does.

I know that the helptext from WRKSYSSTS says "The maximum number of threads that can use the processor concurrently." But I don't quite understand how that could be specific to a memory pool.

Re: the specific situation that brings up this question, we have a rather heavily loaded Tomcat server running on a customer box. And it's running in a 64-bit Java 7 JVM, with max heap of 4G, in a subsystem that has a 6G private pool with Max Active 200.

WRKSYSSTS shows that the JVM job for the Tomcat server has over 700 threads.

Should the Max Active be bumped up? Are there ways, other than trial-and-error, to optimize Max Active?

--
JHHL