Persistent Database Connection Pooling

A persistent database connection pool maintains a set of already-logged-in database connections, and hands them out to applications, on-demand.

Web-based applications, and other transient apps, which connect directly to the database, run queries, and disconnect, often spend more time connecting than actually running queries.

This "connection delay" makes apps feel unresponsive and reduces throughput unless many queries are run per connection.

Applications which connect to a pool, acquire an already-logged in database connection, run queries, and release the connection, generally perform better.

Apps feel responsive and overall throughput doesn't depend nearly as much on the number of queries that are run per connection.

(Current Benchmarks)

Many connection pools run within application servers or web servers, can only be used by applications running within that server, and can usually only by applications written in a particular language. Idle database connections in one application's pool are not typically available to other applications.

Since SQL Relay runs as a standalone process, its database connections are available to all applications, even applications running on other machines, across the network.