Sergey Nivens - Fotolia
Published: 01 May 2015
Think for a moment about the term database management system. It strings together Latin and Greek derivatives. It has eight syllables. Now think about the technology behind the words. It powers operational and analytics applications; it's the very source of business livelihood. For human beings, a comparable resource comes to mind: water. One Anglo-Saxon word, two syllables.
The colorless initialism DBMS also fails to get across the fundamental purpose database systems serve. Their importance should never be understated -- and neither should evaluating the different categories before making a purchase. The wrong choice can leave your data-driven business operations feeling parched.
In this special edition of Business Information, data management consultant and author Craig S. Mullins gives advice on today's database product market. It's not the one of 30, 20, 10 or even five years ago, when the relational database was essentially the only database in town. Today, large-scale Web applications, streaming video and audio information and social media data are too much for conventional relational systems to handle. New data types call for radically new approaches, like storing data in memory instead of on disk, or stepping aside from systems based on the SQL programming language and turning to NoSQL technologies.
Not that relational is going anywhere. In fact, as Mullins advises, "It's wise to evaluate relational database options first." That's because they can be used for so many purposes: data management and analytics applications, online transaction processing and batch processing, to name a few.
But organizations need to follow guidelines when evaluating databases, relational or otherwise. Mullins recommends matching any shopping list against criteria such as the underlying architecture or administration and installation requirements. Then there's the question of whether the job market can provide the types of programming and development skills you need. That becomes a critical one to answer when considering newer, NoSQL databases.
But even if you decide to do things the NoSQL way, which way is that? There are four main product categories, each suited to particular uses. And, Mullins warns, there are no standards, as there are with relational systems. In the end, a cocktail of different database systems may be what your business needs call for: the stalwart consistency of relational combined with the flexibility and brawn of NoSQL.
Other analogies apply, too. Just as the skyscraper relies on the appropriately designed steel frame to reach incredible heights, your data needs the right underlying database management system for support. Only then will it be able to deliver the information and insight that enables you to make better business decisions. And here's another, very human one. It's what allows us our every move: bones. One word, one syllable.