Another cause of "went away" messages is a query that exceeds the
configured memory resources on the server.  Taking a look at MySQL's
logs may shed more light & give hints for which buffers need tuning if
that's the problem.

Yes, with a simple pool event listener you can ensure the liveliness of
connections before the pool hands them out for use.  Usage example is
attached.

Cheers,
Jason

class LookLively(object):
     """Ensures that MySQL connections checked out of the
        pool are alive."""

     def checkout(self, dbapi_con, con_record, con_proxy):
         try:
             try:
                 dbapi_con.ping(False)
             except TypeError:
                 dbapi_con.ping()
         except dbapi_con.OperationalError, ex:
             if ex.args[0] in (2006, 2013, 2014, 2045, 2055):
                 # caught by pool, which will retry with a new connection
                 raise exc.DisconnectionError()
             else:
                 raise

mysql-away.py 2K Download