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.
This is for NB 6.7 RC 3. I am using a quad core system on Windows XP, with Java 1.6.0_13. I have marked this as a P1 because NetBeans freezes and becomes completely unresponsive for me. My only recourse is to kill the app, and sometimes I must delete the cache dir before restarting (or the problem recurs). While using NetBeans, I notice that the CPU often jumps to 25% and stays there for a long time. I suspect this may be related to classpath scanning. While this CPU utilization is occurring, under some circumstances (when? I don't know) the memory consumption will grow quickly to fill the allocated heap space. At that point (with the CPU still pegged at 25% and the heap full), the application will become unresponsive. The NB UI does not respond, and the windows "close" widget in the upper right does not cause the application to close. I am forced to terminate the process. Often when this happens and I terminate the process and restart NetBeans, the heap will grow very quickly (within 1 minute) to the allocated heap size and the same problem occurs. If the problem recurs in this way, I am able to clear the problem by terminating the app and deleting NetBeans' cache dir. I was initially running with the default heap size (NB was taking 512MB). I tweaked the conf file to use 800MB instead, but that didn't relieve the problem -- the heap usage grew to 800MB and the app froze again. My setup: Java Free-Form Project, about 2500 source files (with 84 jar dependencies), about 215 test files (with the same 84 jar dependencies, plus dependency on source files dir). The code is proprietary and can't be shared.
Created attachment 83905 [details] output of jstack
Created attachment 83970 [details] messages.log file
reassign to editor for evaluation
It looks like there was a refactoring task running and it was compiling some java sources. I'm not sure why was it running or for how long. Was there any visual indication of some background activity in the IDE (eg progress bar, hourglass cursor, etc)? Could you also create several successive threaddumps during the IDE activity and attach them here? It would be good to have a reproducible usecase, eg. to know what operation or user action triggers the IDE activity. We could then try to simulate the problem locally on some other project. Thanks
Unfortunately there is no feedback from reported, I'm afraid that without requested info we cannot do anything about it, since we weren't able to reproduce it, Closing as Invalid for now. Feel free to reopen and provide requested info if it is still reproducible thanks
Well for the record, this still occurs for me (with the same frequency, which means occasionally), in 6.7.1. It's a pretty severe problem -- the entire IDE locks up and I have to kill the process and blow away the cache. I just don't have a whole lot of time to help you guys diagnose/reproduce it. I don't really think it makes sense to close a bug that's reproducible, just because *you* can't reproduce it. The issue exists.