On Nov 16, 2007, at 17:38, Chris Bowditch wrote:
<snip />Both the 'theoretical limit' issue and the 'too many threads' should be resolved.Great :) Although whilst testing your latest patch the memory leak seems to be back :(The test has been running for a few hours now and progressing very slowly with the JVM on the limit of the heap. Where as running the test with Jeremias' patch, the memory stays nice and lean well within the max heap and the documents are processed much faster.No OOM Error yet, but I'm sure it can't be far away. I just wanted to update you before I go offline for the weekend.
Can you check the patch in attach? Only a small difference with the previous one.
Tested it here, and I don't see any immediate leakage anymore.One 71-page document, and with -Xmx256M, the heap never exceeded +/-85M. Each document rendered consistently within 3-4s (ran up to 1000 subsequent renderings).
The number of threads never exceeded 9.
propcache.diff
Description: Binary data
Cheers Andreas
