Thanks

For the info and the patch.

I suppose my biggest issue is I don't understand the context Jaybird was working under for the JCA side of things to properly track this down myself.

I'll grab it from CVS, compile it and and run it tonight and see what we get. The thread kill however seems to have limited the damage though so it may no be conclusive without a few days of running.

By the way in answer to a previous question about locking is, is that it is done through the hibernate layer. I think lock for update uses SQL syntax for this.

Holding thumbs for the patch. Will post feedback on how it goes

Andrew