https://bz.apache.org/bugzilla/show_bug.cgi?id=63565
--- Comment #13 from Philippe Mouawad <[email protected]> --- (In reply to Mohamed Ibrahim from comment #10) > Hi Phillipe, > > Sorry It's my mistake.. I apologise, I found the right log. attached it to > the attachments > > > Sharing the code is not an issue. Just a clarification needed from your side. > 1. If our sampler code is leaking, It would have kept the objects for longer > time. If you have gone through the gc log, It would be very clear. There is > no much objects has been promoted to tenure space. > > 2. There would be increase in heap memory for storing referenced objects. We > didn't observe that pattern here in our graph. > > 3. In the Young gen space of 3.x GB, Newly created objects were completely > occupying the space and its getting cleared off every-time when it reaches > the threshold. > > 4. The problem here is for clearing the Objects in the Young gen is kept on > increasing. If you again have a look at gc log, there is a section called > Other which has a section called Ref proc. That is kept on increasing all > over the time during the run. > > > I will share you the code as well at the earliest. Thanks! I didn't go through gc logs yet, I am waiting for : - code - JMX plan if possible or at least run Tools > Schematic view - hprof -- You are receiving this mail because: You are the assignee for the bug.
