Twitter Will Not Use Cassandra To Store Tweets
Twitter announce they will be sticking to their MySQL storage system, but will use Cassandra for geolocation and analytics.
Back in March 2010, it was reported that Twitter was planning to replace their MySQL database system with Cassandra. However, Twitter have now announced that they will not be using Cassandra to store tweets, and will instead maintain their existing MySQL-based storage system.
While this may be MySQL ammunition in the NoSQL/MySQL debate, Twitter have clarified that, although they will not use Cassandra to store tweets, they will use Cassandra in production for geolocation and analytics.
“We’re investing in Cassandra every day. It’ll be with us for a long time and our usage of it will only grow,” reads the blog post.
But why isn’t Twitter backing Cassandra for storing tweets? Todd Hoff has blogged that the answer could be as simple as ‘if it ain’t broke, don’t fix it!’ Writing new code from scratch isn’t always the best business decision. This conservative train of thought could have been been encouraged by the performance problems Twitter has been experiencing recently, including an over-capacity sub-network, which resulted in a slew of fail whales. This is alluded to in Twitter’s blog post, which states “we believe that this isn’t the time to make large scale migration to a new technology.” When it comes to migrating to NoSQL storage, Twitter has other things on its mind.