From: Bill B. <bi...@jb...> - 2003-07-01 14:06:56
|
The metadata stuff would only affect boot time. I'm more concerned about runtime performance and overall memory footprint. Thanks again for your work, Bill > -----Original Message----- > From: jbo...@li... > [mailto:jbo...@li...]On Behalf Of Jon Barnett > Sent: Tuesday, July 01, 2003 8:05 AM > To: jbo...@li... > Subject: RE: [JBoss-user] On the performance trail - again > > > As an addition to this, I should point out that you do see the memory > reclaimed eventually. About three to four hours in idle time on my test > system, sees the RSS drop to around the levels seen with the trimToSize() > addition - probably the time it takes for the GC to find the areas to > reclaim. > > However, this probably detracts from the performance. If I'm reading it > correctly, 20Mb of memory is a lot to clean up, and in my books, a lot to > give up. Since it happens at the metadata-read stage of proceedings, I > guess redeployments are also going to be affected. If you have a big > deployment, you might appreciate the immediate memory reclaim and the > reduced GC interruption later. > > JonB > > > -----Original Message----- > > From: jbo...@li... > > [mailto:jbo...@li...]On Behalf Of Jon Barnett > > Sent: Tuesday, 1 July 2003 11:57 AM > > To: jbo...@li... > > Subject: RE: [JBoss-user] On the performance trail - again > > > > > > As an aside (and up front before the other numbers), I cleaned up some > > code around the MetaData area so that ArrayLists were trimmed and the > > other methods getUniqueChild, etc called my new getChildArrayByTagName > > method. I also finalized methods in this area and Containers and took > out > > variable declarations within loops in these areas. The results were: RT: > > 115 ms (no change), min. 30 ms, max 3405 ms, VSZ: 183516Kb (down 0.37%), > > RSS: 106756Kb (down 17.57%). I probably gave up any performance gains > for > > cleaning memory on the spot. It was interesting the in-memory use > dropped > > that much. But IBM take the opposite view - sacrifice memory for speed. > |