It is a pleasure to introduce guest blogger, Alain MICHEL, business advisor to RDB Consulting.
I remember fifteen years ago, the data world was simple. Part of my job was to deliver productivity web applications to the business. Every time I went to our software engineers, they gave me a choice so to speak: ‘any architecture you want, as long as it is three-tier architecture; and you may select the storage tier from this list: RDBMS.
Fast forward to today: mobility and connectivity are spreading fast and the number of connected devices has risen past 10 Billion. I find myself like millions, lost in cyberspace reviewing concepts on Wikipedia, checking friends on Facebook or buying items on-line.
With the stupendous rise of the real-time Web, Big Data is challenging the core notions of how data is stored and managed. This data tsunami is led by:
- An incredible growth of file volumes, to a point where TB databases are now common
- The shift from old monthly reports to the expectation of real-time information analysis
- The ability to move digital content around faster, leading to applications that access data as it is created
Along with Google, Amazon, Facebook and other major internet companies, surged NoSQL. Their challenges in dealing with data were such that traditional RDBMS solutions could not cope. NoSQL stands for “not only SQL”. I agree this name is not ideal but it has now stuck.
NoSQL is a collection of database management systems, different from the common RDBMS model. Their fundamental distinction is that they do not use SQL as their query language. NoSQL tools focus on the task at hand. Their power comes from their optimization to resolve very well the primary problem at hand.
Who should look at NoSQL? I think of established businesses with systems deployed on relational technology and that are reaching a scalability cost pain-point. And of course multimedia companies for whom Big Data is an opportunity, not a challenge: those organizations whose core business is about collecting, analysing and serving data.
Big Data is all the excitement. Exotic names abound, and experts are still pondering how best to apply these wonderful NoSQL architectures and tools.
- CouchDB, MongoDB, Hadoop, MapReduce…
- Voldemort. Created by LinkedIn and now an open source project
- Cassandra. Started by Facebook, today an Apache top project. Used by Facebook, Twitter, Rackspace… Netflix uses it on the Amazon’s cloud.
Clearly a new era is dawning in the data universe. However until NoSQL shifts from expectations to productivity, my view is that traditional RDBMS will continue to be the platform of choice for typical applications.
Contact RDB Consulting for database management and maintenance – Jennifer Mbesa, firstname.lastname@example.org or +27 (0)11 807 7663