I Want to Have My Data Cake and Eat It Too!

On Point2 minutes readOct 17th, 2013
SHARE +

As the amount of worldwide data has exploded, so have the types of databases, applications and software to help organizations mine this data overload. But how do organizations ensure these new technologies will work seamlessly together when plugged into one platform? An open architecture is the answer. Without this fundamental requirement, organizations will waste resources constantly re-architecting. The Unisys data analytics reference architecture – for example – provides a baseline to adopt big data and Not Only SQL (NoSQL) technologies. It also provides a starting point for the majority of data analytics implementations and helps identify gaps that may exist in a data analytics infrastructure.

The recent rise of big data has produced numerous solutions and frameworks that have traditionally been mainly developer- or programmer-focused. The whole NoSQL movement” has been focused on alternatives to the good old SQL-compliant relational database. So is the relational database management system (RDBMS) now dead with big data? Not at all. More RDBMS and SQL-driven applications exist now than applications on NoSQL databases. This is even more evident in the world of data analytics. It is no wonder projects like Hive and its SQL-like query language HiveQL and Hadapt try to bring SQL back into the NoSQL world. These efforts are a realization of the undeniable fact: Many customers want to work with data in a multitude of different ways.

With many NoSQL databases now providing SQL support, will there be one database to rule them all? Not anytime soon. The number of different solutions out there and the degree of specialization is still far too great. Will there be a convergence around platforms and standards? Yes, but it will take a while given the requirements. Good luck finding a techie that knows Hadoop, SQL, HiveQL, PIG, CQL, SPARQL, XQuery, REST, and all the other standard, proprietary, and API-based ways to query a NoSQL database. As an analytic creation moves away from the developers – back into the hands of the data analysts – vendors and developers are reworking their database engines to be more SQL-friendly and serve this very different customer base.Developing a reference architecture now – with tools that are a natural fit for your internal data customers – is critical. The Unisys data analytics reference architecture provides a baseline to adopt big data and NoSQL technologies with both the developer and the data analyst in mind.

Tags-   Big data NoSQL technologies