This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.

Bug 101895

Summary:	to look into peak memory usage during reverse engineering
Product:	uml	Reporter:	Peter Lam <petersl>
Component:	Reverse Engineering	Assignee:	issues@uml <issues>
Status:	NEW ---
Severity:	blocker	CC:	issues
Priority:	P3	Keywords:	PERFORMANCE
Version:	5.x
Hardware:	All
OS:	All
Issue Type:	TASK	Exception Reporter:
Bug Depends on:
Bug Blocks:	103340

Description Peter Lam 2007-04-20 06:28:54 UTC

After reverse engineering (RE) the uml core sources (~2700+ classes) on my
Windows XP laptop (1.8GHz, 1GB RAM) and it completed in less than 5 minutes
which is good. However, there's a problem with memory usage. After the RE
action, the memory usage is 599208K from the Task Manager and it never came
down. The memory usage should be released or GC'ed after the action is
completed. Because of memory was not released, ide performance decreases.

Comment 1 Viktor Lapitski 2007-04-20 06:55:22 UTC

looks very similar to IZ 100257

Comment 2 Viktor Lapitski 2007-04-23 08:26:45 UTC

We dont' have an explicit way to return back to the OS the unused memory 
taken by JVM.
During reverse engineering the heap memory usage is going up to more than 400+M 
what is normal to result in 600M or more of RAM taken by JVM. After the RE the
heap memory usage is down to about 120+M (can be observed by explicitly running
garbage collecting by clicking on memory monitor - it will result in 300+M of
free heap). 
Thus we can see that the memory is released after the action is completed. That
effectively renders the originally stated problem to be "not reproducible".
Though it still makes sense to look into peaked memory usage during the RE, so i
leave the bug open as a placeholder for it.