[jboss-dev-forums] [Design the new POJO MicroContainer] - Locking and the DeployersImpl.shutdown()

[adrian@jboss.org]

Related to this:
http://jira.jboss.com/jira/browse/JBDEPLOY-19

There's some code/test I'm going break
relating to shutdown and single deployment.

It was DeployerSingleDeploymentTestCase::testMultiThreadsAndShutdown()

This is incorrect behaviour.

The shutdown should not be blocked by a deployment, it were
it could "deadlock" the shutdown when the deployment is misbehaving (e.g. a broken service
that is waiting on a socket forever because of a network split)

I'm going to remove this test. 

The test actually suffers from a race anyway.
The deployment threads (as written currently) could run after the shutdown thread
gets the lock which means the test would fail.

View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4137759#4137759

Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4137759