Noel J. Bergman wrote:
> Stefano Bagnara wrote:
> > let's collect some proof that the jvm problem is correctly workarounded
> > this way, then we'll decide how to fix it.
> What do you think that I've been doing? FWIW, heap dumps are in
> ~noel/JAMES-only.
FWIW, now that we are looking for this specific issue, you can spot it in
the old heap dump from the summer that is still in that directory, as well
as clearly see it in the 2.3.0 log, and the effect of the change in the
2.3.1-dev heap dump.
Two things that helped make the heap dump analysis easier were lowering the
size allowed for dnsjava's cache, and watching the heap stats so that I dump
the heap immediately after a full GC (the first SITES dump in the 2.3.1-dev
shows how much harder it is to analyze if you DON'T do it immediately after
a full GC).
--- Noel
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
