From: <glassfish_at_javadesktop.org>
Date: Tue, 17 Nov 2009 04:16:15 PST

Hi All,
We are facing similar incidence in our clustered environment where few of the nodes go no responsive and when kill -3 is run in the JVM logs we can see the following
<writer thread='1731144608'/>
<dependency_failed type='unique_concrete_method' x='java/util/ArrayList add (Ljava/lang/Object;)Z' witness='javax/management/AttributeList' stamp='0.672'/>
<writer thread='1707985824'/>
<dependency_failed type='abstract_with_unique_concrete_subtype' ctxk='java/nio/ByteBuffer' x='java/nio/HeapByteBuffer' witness='java/nio/DirectByteBuffer' stamp='3.519'/>
<dependency_failed type='abstract_with_unique_concrete_subtype' ctxk='java/nio/ByteBuffer' x='java/nio/HeapByteBuffer' witness='java/nio/DirectByteBuffer' stamp='3.519'/>
<dependency_failed type='abstract_with_unique_concrete_subtype' ctxk='java/nio/ByteBuffer' x='java/nio/HeapByteBuffer' witness='java/nio/DirectByteBuffer' stamp='3.519'/>

Ours is a 64 bit environment as well..

The nodes goes down every odd day...

any response/resolution on this is greatly welcome..

Thanks

Ravi
[Message sent by forum member 'vtrravikumar' ]

http://forums.java.net/jive/thread.jspa?messageID=372135