I have a general question about James' jdbc connectivity. How efficient
is it? What I mean is does it employ connection pooling? I was looking
through the code and it seemed to suggest that it is done in the avalon
framework, but reading the docs in config.xml, I began to second guess
it. Just to clarify that I my idea of pooling is clear, let me explain
some. Typically in a high volume web application, one would create a
database pool of n connections that are opened when the application is
started and kept open for queries. Built in functionality expires these
connections, grows the pool, and shrinks it as necessary. After leaving
weblogic many years ago, I have implemented my own Connection pooling
library that does this similarly and had planned on using it as a
wrapper to the mysql driver for James, but my question is do I need to
do this? Does Avalon or James already do this?
What leads me to really be thinking about this is the
JDBCVirtualUserTable class (yeah, I know enough with the virtual hosting
stuff). If this class is constantly querying the James database it is
going to put significant additional load if it is constantly opening and
closing connections.
Bottom line. Should I use my connection pooling or is it redundant?
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
