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.
Scanning (precisely JMManager.cleanUpCodebases) is running forever and it loads processor for 100%. Also NB memory meter is going up to 50 MB then down to 25 MB and so on. I have opened the same module from diffrent branches. Nothing else. To reproduce: - start IDE woth fresh userdir - open nb_all/ide module from trunk - open nb_all/ide module from release55 branch. Progress bar "Scanning Project Classpath" never finishes and it overloads processor forever. Build 20060224-1109, JDK1.5.0_06, WindowsXP.
Created attachment 29007 [details] Thread dump
Created attachment 29008 [details] Screen shot.
Is this reproducible?
Yes, it is with build 20060331-0804. Progress bar disappeared but CPU is at 100% and memory meter goes up and down continuously.
It is really annoying and I am able to easily reproduce it in NB55: - start IDE - open nb_all/openide project from release55 branch (just this project, not required projects) - it starts to scan classpath. When it is finished (progress bar disappears), processor gets to 100% and memory meter goes up and down. It is because JMManager.cleanUpCodebases is running forever. Build NB55 Beta2 200607170300, JDK1.5.0_06, WindowsXP.
Created attachment 31942 [details] Thread dump from NB55
I cannot reproduce it on my machine.
cannot reproduce either, Jirko can you try it again, please?
Guys, how many times should I try it to convince you that it really behaves like I described? If you don't believe, please come and look at my display. Last time I reproduced it with build 20060817-0525 NB55, WindowsXP, JDK1.6.0-b93.
Jirko, the problem is probably specific to your machine. None else (afaik) has such problems. I don't depreciate your report, I just need more information. There are hundreds of developers, who can open openide without any problem. We appreciate, if you can: 1. Reproduce this issue one other machine than yours 2. tell us, if you use soft links or similar stuff. Thanks
You are right I can't reproduce it on other computer. If I find a reason, why it doesn't work on my computer, I will reopen it.
Verified.
I might have a hint on analysing this (besides directly attaching debugger to Jirka's machine, which seems most strightforward if Jirka can reproduce it): Take a heap dump (jmap -dump:file) and analyse the data structures the JMManager sees (it _should_ be possible to get even to the stack-referenced structures). That might shed some light into why JMManager isn't able to finish the scan.
You are out of luck :-( I am not able to reproduce it anymore.
Reorganization of java component