About caching and scalability

Explanation of LiveWhale’s cache layer.

LiveWhale is finely tuned for speed using performance analysis tools like XDebug and benchmarking tools like Apache Benchmark. It makes use of a special cache layer which is designed to achieve high performance without sacrificing the functionality that older caching models do. In addition to this, LiveWhale makes use of strong client side browser caching to create the lowest possible load on your server.

For page caching, LiveWhale uses neither the push nor pull model in the strict sense. In effect, it follows a middle path, drawing from the benefits of these models while trying to avoid their limitations and flaws. Our CMS is able to do this by providing intelligent caching solutions to deal with the many shortcomings of both the classic push and pull models.

Adopting methods used by sites like Facebook and Yahoo, LiveWhale’s cache layer acts as a buffer between your web pages and the CMS itself. This means that if the CMS goes down or a resource, such as a database, becomes unavailable, your web site will continue to serve cached content just fine. If this happens to be the case, LiveWhale will quietly log the issue in the server log. The cache layer will also talk to the CMS and make sure that fresh content is passively and continually pushed so that it remains up to date for your web server to serve, and users never have to worry about publishing their content manually.

Aside from page caching, strong caching techniques are used throughout the site. For example, if an RSS feed is embedded in your web page, that feed is cached locally and will not have to be requested or parsed again on subsequent requests. Widgets themselves are also cached, and do not expire until events in the system trigger their expiration, such as changes to content that may specifically effect that particular widget’s output. Cache layers like this are built into all major components of the CMS.