NoSQL - Factors Supporting the Adoption of Non-Relational Databases

Abstract

Relational databases have been around for decades and are still in use for general data storage needs. The web has created usage patterns for data storage and querying where current implementations of relational databases fit poorly. NoSQL is an umbrella term for various new data stores which emerged virtually simultaneously at the time when relational databases were the de facto standard for data storage. It is claimed that the new data stores address the changed needs better than the relational databases. The simple reason behind this phenomenon is the cost. If the systems are too slow or can't handle the load, the users will go to a competing site, and continue spending their time there watching 'wrong' advertisements. On the other hand, scaling relational databases is hard. It can be done and commercial RDBMS vendors have such systems available but it is out of reach of a startup because of the price tag included. This study reveals the reasons why many companies have found existing data storage solutions inadequate and developed new data stores.

Asiasanat:databases, database scalability, data models, nosql