The concept of caching is simple, in that it provides pre-processed result sets for oft-used queries (often referred to as static data). These result sets can be extracted from the database and be provided to the end user in some alternate manner. In principle:

• Eliminating repeated requests for static data reduces load on the the server and allows it to serve transactional data more efficiently.

However, in practice, improvements in database servers, disk drives and memory have made it difficult to obtain improvements in performance unless:

• The performance of the underlying system is extremely poor, for example in stressed, resource limited systems.
• All incremental improvements are considered appropriate, for example in high-volume OLTP environments.

In this situation, each of the above situations can be considered as a different configuration issue.

ADOdb provides 2 distinct solutions to the underlying configuration:

1. A remote memory caching server based on Memcached, using either the memcache or memcached clients.
2. A local memory caching service based on APCu

Once implemented, the coding methodologies are the same.