Connection pooling significantly reduces the overhead associated with creating and destroying database connections. Connection pooling also improves the scalability of solutions by reducing the number of concurrent connections a database must maintain-many of which sit idle for a significant portion of their lifetimes. With connection pooling, instead of creating and opening a new connection object whenever you need one, you take an already open connection from the connection pool. When you have finished using the connection, instead of closing it, you return it to the pool and allow other code to use it.