For what it is worth, I have had such problems before.

In my case, the host I was trying to reach had closed down the server
because xinetd had seen too many requests in too short a time. The
immediate problem was only solved when the daemon had waited 30 minutes
to re-open and allow new requests.

The solution to the problem was increasing the number of allowed
connections for firebird - done in the firebird service module
per_source = 200

Of course, you may not be running linux or mac-osx - so this may be
useless advice. I don't know how to open up the number elsewhere.


And oh yes, the number being requested was a bug due to tomcat
restarting stale (dead) code as it thought it was doing me a favor with
serialization.