Re: [JAVA2D] sun.java2d.Disposer questions (possible leak)

2009-03-15 Thread java2d
I am experiencing the same problem (environment: tomcat). I choose to initialize Java2D early during boot of the application server instead having a memory later when it is being initialized by an app. This class is called as a wrapper for the original tomcat bootstrap class. This is only a

Re: [JAVA2D] sun.java2d.Disposer questions (possible leak)

2007-01-26 Thread java2d
Is there any chance of getting a fix for this bug in the somewhat near future? :-) [Message sent by forum member 'plethora' (plethora)] http://forums.java.net/jive/thread.jspa?messageID=198778 === To unsubscribe, send email

Re: [JAVA2D] sun.java2d.Disposer questions (possible leak)

2007-01-26 Thread Dmitri Trembovetski
I've filed this bug a while ago: 6489540: The Disposer thread could cause memory leaks in user applications http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6489540 Unfortunately without some changes in another area it can not be fully fixed, so I filed this bug: 6501120:

Re: [JAVA2D] sun.java2d.Disposer questions (possible leak)

2007-01-26 Thread java2d
Dmitri, thanks for the detailed update. [Message sent by forum member 'plethora' (plethora)] http://forums.java.net/jive/thread.jspa?messageID=198831 === To unsubscribe, send email to [EMAIL PROTECTED] and include in the

Re: [JAVA2D] sun.java2d.Disposer questions (possible leak)

2006-11-03 Thread java2d
Dmitri, Thanks for the quick response and for entering the bug. I will take a look once it becomes visible. It might be a good idea to inspect all active (threaded) run-until-JVM-shutdown subsystems for contextClassLoader reference leaks. Regards, Taras [Message sent by forum member 'plethora'

Re: [JAVA2D] sun.java2d.Disposer questions (possible leak)

2006-11-02 Thread java2d
OK, I've filed a bug (will appear on bug parade in a day): 6489540: The Disposer thread could cause memory leaks in user applications I've also found out that there's no way to address the first problem about inheritable thread local storage (see bug's evaluation). But the second one - with

Re: [JAVA2D] sun.java2d.Disposer questions (possible leak)

2006-06-27 Thread Dmitri Trembovetski
On Mon, Jun 26, 2006 at 10:47:29PM -0700, [EMAIL PROTECTED] wrote: Sure, let's say I manageg to remove all the inheritableThreadLocals. The next issue is the contextClassLoader that is propagated to the Java2D Disposer Thread. Which, of course, is exactly the ClassLoader I'd like to

Re: [JAVA2D] sun.java2d.Disposer questions (possible leak)

2006-06-27 Thread Dmitri Trembovetski
But before I start asking I need to understand more about this. Do you set your own classloader, or is it the default one? Better yet, is there any way you could create a small test case illustrating the problem? Thanks, Dmitri On Tue, Jun 27, 2006 at 06:18:19AM -0700, Dmitri

Re: [JAVA2D] sun.java2d.Disposer questions (possible leak)

2006-06-26 Thread Dmitri Trembovetski
Hi, On Sun, Jun 25, 2006 at 01:02:35PM -0700, [EMAIL PROTECTED] wrote: Hi Dmitri, Is Mustang b89 recent enough to test? Yes, that should do it. Out of curiosity, what was changed wrt the Disposer in Mustang? Is the disposer thread initialized on JVM startup? The change is

Re: [JAVA2D] sun.java2d.Disposer questions (possible leak)

2006-06-26 Thread java2d
The leak is still there in Mustang b89. [Message sent by forum member 'plethora' (plethora)] http://forums.java.net/jive/thread.jspa?messageID=126968 === To unsubscribe, send email to [EMAIL PROTECTED] and include in the

Re: [JAVA2D] sun.java2d.Disposer questions (possible leak)

2006-06-26 Thread Dmitri Trembovetski
Could you try the workaround I've suggested? (do some java2d activity - in particular, create a BufferedImage and render to it - on a thread created either prior to creating and setting the InheritableThreadLocalStorage or started from a separate thread). I'll think on what we can do

Re: [JAVA2D] sun.java2d.Disposer questions (possible leak)

2006-06-26 Thread Dmitri Trembovetski
Hi, On Mon, Jun 26, 2006 at 07:22:35AM -0700, Dmitri Trembovetski wrote: Could you try the workaround I've suggested? (do some java2d activity - in particular, create a BufferedImage and render to it - on a thread created either prior to creating and setting the

Re: [JAVA2D] sun.java2d.Disposer questions (possible leak)

2006-06-26 Thread java2d
Sure, let's say I manageg to remove all the inheritableThreadLocals. The next issue is the contextClassLoader that is propagated to the Java2D Disposer Thread. Which, of course, is exactly the ClassLoader I'd like to unload (the one with 1GB+ data referenced). Any idea how to do that? [Message

Re: [JAVA2D] sun.java2d.Disposer questions (possible leak)

2006-06-25 Thread java2d
Hi Dmitri, Is Mustang b89 recent enough to test? Out of curiosity, what was changed wrt the Disposer in Mustang? Is the disposer thread initialized on JVM startup? BTW, the application is a J2EE web app, sometimes serving some graphs with historical data. The InheritableThreadLocal comes from

Re: [JAVA2D] sun.java2d.Disposer questions (possible leak)

2006-06-12 Thread Dmitri Trembovetski
Hello, On Fri, Jun 09, 2006 at 01:52:32PM -0700, [EMAIL PROTECTED] wrote: First, is there a good description available on how the sun.java2d.Disposer class works in Java 1.5.0? There isn't such a description. Disoser is an internal mechanism for disposing of graphics-related resources