Campfire #3 : NoSQL Databases - NoSQL or Not only SQL? [Webinar 41min]

On February 14, 2013, Mertech continued the Campfire initiative with a session on NoSQL databases. Mertech representatives Oliver Nelson (Senior Developer) and Anton Voskresenskiy (Developer) led the discussion. If you missed it, the session is available above.

Presented by Oliver, mainly focused on discussing the main characteristics of NoSQL databases, the different types, major players, and why using a NoSQL database makes sense. In a nutshell, NoSQL differs from the standard SQL relationship structure which can limit speed and scalability for large data sets. The approach is to use a key-value store instead of a relational structure, which enables NoSQL’s biggest advantage, distribution: the ability to scale up quickly with superior performance. There are many players in the space including Oracle.

NoSQL databases are best suited to use cases which require the storage and retrieval of large amounts of data. For example, logging of clicks on a website results in significant amounts of data, which don’t need to use a relationship structure. Some examples of well-known companies that have implemented NoSQL include Craigslist, who implemented MongoDB for archived postings, but still maintains MySQL for active postings, and Facebook, who created Cassandra when they expanded their messaging service to include all types of messages including email and text.

The conversation then moved to Anton, who talked about NoSQL database variations. From triple stores, which are well suited for dynamic relational data, to document stores (for storing XML, JSON, etc.) and graphs, which focus on deeply related data.

Other discussion points were brought up, including whether it makes sense to use NoSQL for e-commerce. MongoDB is typically marketed as a great fit for e-commerce and document processing systems, and the nature of the data needed in managing an e-commerce website (tracking clicks, etc.) points to using NoSQL. Other uses in all industries might include software testing, market research, or employee productivity measurement, which all contain large volumes of unstructured data.

All in all, it was an informative session, and we look forward to future events.