> > 1. With .NET you have a connection string in each piece of code using > the pool instead of instantiating a connection to an existing pool. > 2. If you fail to leave at least 1 .NET connection open, the pool goes > away, and needs to be recreated the next time it is needed, where a ds > would stay available for fast connectivity-even on the first > connection to the pool. > 3. You need to have precisely the same connection string in the app, > where a reference to a ds class is all that is needed for a java app > server. >
Pretty much spot on (from my understanding of connection pooling in Java). In .NET, the pooling is a bit more transparent, other than modifying your connection strings (min/max pool size, timeout etc) you don't really need to do anything to use them. -Jamie
