ApacheCon Session
Load-balancing with Apache HTTPD 2.2 and later
When operating modern, high-profile websites based on J2EE or one of the various web frameworks like Rails or Django, a single machine is rarely enough to handle the demand. It usually requires a whole farm of dedicated application servers load-balanced by one or more frontend servers. Learn how to leverage the Apache HTTP server to implement a robust, secure and effective load-balancer to improve response times and to provide automatic failover for your critical services. Topics covered in this session include: connecting backend servers via HTTP or AJP, using several backend clusters by partioning the URL space, using rewrite rules to allow flexible request handling, using sticky sessions to provide persistence, using special headers to communicate with the backends, offloading SSL and filtering content, operating multiple frontends in one or more physical locations as well as general load-balancing algorithms, common caching strategies and browser-based management of the balancers. Some useful tips and a discussion of the most common pitfalls, based on real-world experience, will form the conclusion of the session.
