maximum memory for gcj-compiled executable?

Lars Andersen lars@rimfaxe.com
Thu Apr 3 14:03:00 GMT 2003


On Thu, 2003-04-03 at 15:17, Ranjit Mathew wrote:
> > > On linux it is about 1kb per http request processed, on win32 it's
> > > 48KB(!) per request!
> > > 
> > > heap size remains the same.
> > > 
> > > Why is it eating up my memory??? It must be something.
> > > 
> > 
> > Coul it be thread spawning?
> > If so, why the difference between win32 and linux
> 
> Is a new Thread created per request? Have you tried
> using a Thread Pool as in, for example, Doug Lea's
> "util.concurrent" package? Do these Threads never
> get destroyed?
> 
> Ranjit.

I am using a thread pool. I spawn 40 threads on startup. And I have just
checked that they actually gets reused properly, and it all seems fine.

Also, there's growth when the program makes a http request itself, and
here there is definitely no thread spawned. But i won't rule it out as
the explanation quite yet, theres a lot going on at the same time.

I'm trying to create a small test program demonstrating the issue for
you all.

Hang on.

/Lars Andersen



More information about the Java mailing list