Using Redis as a Cache Backend in Magento
By: Alexey Samorukov, Aleksandr Lozhechnik, Kirill Morozov | August 12th, 2013
The most common cache storage recommendation for Magento users running multiple web nodes
is to implement the TwoLevels backend – that is, to use Memcache together with the
database. However, there are notable issues with this approach. First, the core_cache_tag
table constantly grows. On average, each web site has about 5 million records. If a system
has multiple web sites and web stores with large catalogs, it can easily grow to 15 million
records in less than a day. Second, the TwoLevels backend is more difficult to maintain
because two services are required to make it work which makes it difficult to analyze cache
content when necessary. Further, memcached itself has limitations such as a maximum object
size and fixed bucket sizes which also contribute to difficult maintenance. Finally, the
TwoLevels backend does not scale well, since using the database as part of the cache
backend adds additional load to the master database server. Additionally, there is no
reliable method for memcached replication.
Download the article to the right to learn more about how using Redis can help.