Ditching the database

Each introduction of the grandest, biggest and latest database comes with a promise in the line of “this database will make all the data available to all your users in a flash”.

Brilliant advances in technology ranging from in-memory, ACID-compliant databases, HADOOP, Hana, super-fast, petabyte-handling databases do provide grounds for database administrators becoming hugely excited. Yet in the BI world, as the user dissatisfaction grows at the same pace as data volumes, this mirage rapidly fades.

When data users start to express frustration of not being able to find answers to their business questions or worse: finding inaccurate and conflicting results, it is time to go back to basics. No super-powered database will solve this problem.

Trading business requirements against technological advances of the database is a recipe for disaster. Understanding business requirements, collecting, cleaning, restructuring and storing relevant data in the database ready for any data enquiry is the mainstay for all Data Warehouse and/or Business Intelligence projects.

When do you ditch the database? Only if it does NOT store relevant business data and if it does not have the ability to supply accurate data in a fast, effective and easy way while being adaptive and resilient enough to weather ever changing enquiry patterns. In such case you may have to revisit the business requirements and make sure that the underlying data structures do support these needs.

Enroll NOW for the Data Warehouse Concepts and Dimensional Modeling Techniques that teaches sound principles to be applied no matter what technology is used.

By Susan Andrè

Other articles to read
Toss the Technology Part 1
Toss the Technology Part 2
Time to Expel Excel

More about Alicornio Africa