Ted Hulick [
http://community.jboss.org/people/Ted.Hulick%40hp.com] created the discussion
"Re: More on BaseClassLoader conflict with Instrumentation...."
To view the discussion, visit:
http://community.jboss.org/message/589248#589248
--------------------------------------------------------------
First, I really appreciate your help...
Ok - I don't control the WAR file...essentially, our product montiors JBoss app
server...the WAR in this case was written by QA...and they included jms.jar in their
deployment. So, you are saying that if they DID NOT include it - it would have found the
one I attached to the base system classloader?
And you are saying that by default the WAR will indeed load it's jms.jar and NOT look
downstream to my jms.jar? Is there a way to override this behavior just for that 1 jar?
Maybe via a property that was set? Can it be done programmatically? We can't ask the
customer to change
their deployments.
On the jboss-classloading.xml to parent first...would this override the WAR file
self-first? Again, can something be done via property or programmatically so customers
don't have to alter their deployment.
Essentially, I'm just added some jars to the classpath using the addURL (and change
protection) on the system classloader and then referring
to those classes in instrumentation....because they make reference to jms classes - I have
to make sure jms.jar is in the system classloader
path as well. When anything else in the app loader loads jms.jar (which is only
interfaces anyhow) classes from their own loader and now
delegate downstream to mine - then Class Cast issues occur as of course, a class instance
in 2 different loaders is not considered to be
the same class.
Again - any help/suggestions is appreciated....
--------------------------------------------------------------
Reply to this message by going to Community
[
http://community.jboss.org/message/589248#589248]
Start a new discussion in JBoss Microcontainer at Community
[
http://community.jboss.org/choose-container!input.jspa?contentType=1&...]