5 Predictions on the Future of Databases

Article Research week 5

Jeff Dozier

POS 410

01 19 15

Kevin Judge

5 predictions on the future of databases

“Michael Stonebraker is a lightning rod in the database world” (5 predictions).  He is very influential in the database world and has helped create many of the systems we use today.  He has more recently worked on VoltDB, on in-memory OLTP system that is much faster than legacy options.  Mr. Stonebraker comes out with his yearly predictions, which always create a buzz and will almost certainly create differences of opinion by many in the industry.  This year, he has given us 5 predictions for the future of the data management industry.  

Oracle University

The highest listed certification (or at least it is listed as number one out of the top five) is for Oracle, and it is considered the number one database platform (ICCP). The Oracle certifications are tightly regimented and have different tracks for database administration, application development, and implementation (ICCP).  If you visit the Oracle University website there are a large number of training variables. The Oracle training requires candidates to attend Oracle training classes prior to testing, and Oracle certification is true ladder in which you have to advance each stage before going to the next (ICCP).  More than 80% of people who earned an Oracle certification reported a salary increase, promotion or career improvement (Oracle Certifications).

One size will fit none

There is increasing evidence, and need for, several types of non-relational database management systems on the horizon.  Per Stonbraker, many vertical markets are looking into alternate methods to using RDBMS rather that legacy RDBMS.  For example there are new developments for columnar architectures for analytics, in-memory architectures for transactions, and NoSQL architectures for simple key-value operations and new data types, as well as graph database architectures.  

Room for competition

There will be four, five, or six categories of database architectures and two to three successful vendors of each (5 predictions). The core of traditional RDBMS will lose popularity and market share in the future, and the RDBMS universe will shrink.  

No SQL makes a comeback

Stonbraker predicts that NoSQL will make a revision and come back with different functionality since the perceived value of purely low level language is all but gone.  He predicts it will go the way of Googles ‘Spanner’ database and rely heavily on ACID (Atomicity, Consistency, Isolation, Durability) properties in future NoSQL iterations.