Analysis & Discussion of Results

From the benchmark results we can see a 42.5% improvement in both number of hits and requests per second when using SQLCache. What is more, the number of connections made by the web server to the back end SQL server has been dramatically reduced from 52 to 1. Furthermore, this single connection was made only once to populate SQLCache's cache.

SQLCache's impact upon your web sites performance and scalability is greatest when your systems are under most load. One of the biggest performance hits to using database servers as repositories of dynamic content is the wait time associated with the "round trip" required to get the data. This wait time includes elements such as the network latency between the machines and importantly the time taken by the database server to process the command and return the results. We can see two potential bottlenecks that will become increasingly severe as your front-end web servers make more connections. Networking bottlenecks can be overcome relatively cheaply, but to increase the capability of your database servers is expensive. Generally these are powerful and expensive machines that will require clustering for high availability, so the cost to upgrade can be prohibitive. SQLCache helps out most here because it reduces the load placed on these types of machines by caching frequently accessed data on the web server. This eliminates the need for the "round trip" saving your network and database resources. In totality SQLCache enables you to serve more content, quicker, using fewer resources.