Brian Swan has another post in his series looking at using the SQL Server driver in PHP applications. This part of the series looks specifically at using the connection pooling option to help improve overall performance of the application.

One topic that came up yesterday [during Jump In! Camp] was connection pooling and how much difference using pooled connections makes in app performance, so that's what I'll look at in this post.

He starts by defining (from Wikipedia) what connection pooling is and then puts it in the context of SQL Server by talking about the handling of connections returned to the pool. He shows how to enable and disable the feature via a parameter in the connection options and shows how to configure the number of seconds a connection stays active in the pool.