I recently added the locks and latches plugin (version 0.4) to hudson (version 1.306) so that I could have more than one executor, but not
have two jobs access the same test database at once. This has been working reasonably well, except for the problems noted by the plugin
author to do with the fact that the job "starts" before checking if the resources are available.

I've now seen another problem. I have only a single job running, but it is stuck at the very start with the following output in the log:

A SCM change trigger started this job
Building remotely on

When I kill the job, I get the output below in hudson.err.log, which I suspect means that a lock is held (perhaps by another thread,
perhaps due to failure to clean up). Because the master is the executor, I can't get System Info the way that I can from slaves; that
would help me see if there is a problematic thread.

INFO: job-name #604 aborted
java.lang.InterruptedException
at java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireInterruptibly(AbstractQueuedSynchronizer.java:813)
at java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireInterruptibly(AbstractQueuedSynchronizer.java:1137)
at java.util.concurrent.locks.ReentrantLock.lockInterruptibly(ReentrantLock.java:312)
at hudson.model.AbstractProject.checkout(AbstractProject.java:823)
at hudson.model.AbstractBuild$AbstractRunner.checkout(AbstractBuild.java:314)
at hudson.model.AbstractBuild$AbstractRunner.run(AbstractBuild.java:266)
at hudson.model.Run.run(Run.java:927)
at hudson.maven.MavenModuleSetBuild.run(MavenModuleSetBuild.java:234)
at hudson.model.ResourceController.execute(ResourceController.java:93)
at hudson.model.Executor.run(Executor.java:119)
May 26, 2009 1:15:55 PM hudson.model.Run run
INFO: job-name #604 main build action completed: FAILURE