Re: [jboss-dev] org.jboss.util.threadpool.BasicThreadPool issue
Sorry to have pointed at you. Scott has a nice comment also:
-------
/**
* A basic thread pool.
* TODO: this port to jdk concurrent still needs to be tested.
*
* @author <a href="mailto:adrian@jboss.org">Adrian Brock</a>
* @author Scott.Stark(a)jboss.org
* @version $Revision$
*/
-------
A combination of head, tail, cat, vi were needed to see through the maze
of the logging - 900MB of server.log is not easy to deal with. :)
So should I make the fix to this class and do a snapshot release for
common-core? What should it be 2.2.2-SNAPSHOT?
Adrian wrote:
I'm not making use of anything, Scott wrote this code. :-)
The issue looks to be related to the conversion
of the code from using Heap (which returns null when empty)
to PriorityQueue.
http://viewvc.jboss.org/cgi-bin/viewvc.cgi/common/common-core/trunk/src/m...
Contrary to your comment, PriorityQueue throws an Exception
when empty on the remove() as opposed to remove(Object):
http://java.sun.com/j2se/1.5.0/docs/api/java/util/AbstractQueue.html#remo...
A simple check for isEmpty() and return null
(assuming the code is already thread safe) should stop the logging?
On Tue, 2007-09-11 at 09:57 -0500, Anil Saldhana wrote:
> Adrian,
> with some progressive changes in my AS5 trunk workspace, while
> running one of the tests in the test suite, the server log is filling up
> with the following exception (it is reaching something like 800MB in log
> size)
> =================
> 2007-09-11 09:40:10,843 DEBUG
> [org.jboss.util.threadpool.BasicThreadPool] Timeou
> t monitor saw unexpected error
> java.util.NoSuchElementException
> at java.util.AbstractQueue.remove(AbstractQueue.java:80)
> at
> org.jboss.util.threadpool.BasicThreadPool.getNextTimeout(BasicThreadPool.java:475)
> at
>
org.jboss.util.threadpool.BasicThreadPool$TimeoutMonitor.run(BasicThreadPool.java:593)
> at java.lang.Thread.run(Thread.java:595)
>
> =====================
>
> I am still checking why this issue is happening. But I see that you are
> making use of a PriorityQueue and whose remove is supposed to be safe
> from any exception.
>
> Regards,
> Anil
>