Description of problem:

I don't know if this is related to my SQL Server jdbc driver or if it 
applies to Oracle and PostgreSQL as well but the Indexer, 
RequestManager, SimpleQueueManager, and DigestQueueManager threads 
die when my database goes down.  These threads have a try catch block 
with a finally block that either call txn.commitTxn() or txn.abortTxn
().  The problem is that when the database goes down a "Connection 
reset by peer: socket write error" exception is thrown by my SQL 
Server jdbc implimentation.  After the exception is thrown the 
finally block is called with commit or abort.  The commit or abort 
then throw another "Connection reset by peer: socket write error" 
which is not caught causing the thread to terminate.