Whoa, I've never heard of this! I mean, I knew that folks in the
Java community highly encouraged using a type 4 JDBC driver when available, but I didn't know the bridge was that bad. We've been using it for well over a year with our web app because this vendor supposedly only has an ODBC driver. The whole reason we're moving our db server to a Linux box is because they finally admitted that the Windows version of their ODBC driver has always had problems. We often have to restart Tomcat and/or the ODBC driver to get our web app talking to the database again, and this can sometimes happen several times during the day. I'm not sure if this is considered a heavy load for a single pair of servers (web server and db server), but our web app has about 2000+ concurrent users at peak times at various points throughout the day.
However, my question is--do you think it's the JDBC-ODBC bridge that's causing these problems? I was new at this when I got here, but the original developer who is no longer with us was always convinced it was their driver, not our textbook connection pooling code or anything like that (I heard he even dropped in the Apache group's connection pooling package in place of his to make sure). However, if Sun's bridge is actually the problem...that would be very interesting. What do you think?
[ August 12, 2004: Message edited by: Stephen Huey ]