I understand the high level concept of connection pooling. What I'm trying to get is a more detailed understanding. In particular, suppose I were to write my own connection pool (in Java of course;) ) . I understand the notion of caching the connections. What i'm not getting is this:
If I have code that calls the connection pooling object or objects, then won't all those connections be created at the time the Connection Pool object is being created (assuming it does this when the constructor runs).
I would think you would have to have a Connection pooling "Server" situation where the Connection pool Object(s) are listening for a connection request. If this is the case, when speaking of creating a connection pool, am I really talking about implementing a special kind of server?