PostgreSQL ramps up its NoSQL game

New developer kit for PostgreSQL eases access to the open source database's recently introduced NoSQL features

NoSQL databases like MongoDB or CouchDB generally compete with each other, but now a relational database is positioning itself as a NoSQL player.

The PostgreSQL Project, which EnterpriseDB supports, added NoSQL-style JSON processing features back in 2012. Now, the company is encouraging further work around that feature set by providing a developer kit to make it easier for programmers to leverage PostgreSQL's JSON functions and build applications around them.

The PGXDK (Postgres Extended Datatype Developer Kit) is designed to allow developers "to use Postgres for the kinds of applications that until recently required a specialized NoSQL-only solution," as EnterpriseDB describes it. A sample application is also included to make it easier for developers to get a leg up on working with the product. The whole package will be made available through AWS as a machine image (PostgreSQL has long been a staple Amazon offering).

A few key rationalizations exist for adding NoSQL functionality to columnar databases. For one, it allows NoSQL work to be processed with the same atomicity and consistency as a conventional database, and it allows developers to use conventional SQL instead of NoSQL's own querying system.

Another is performance, as EnterpriseDB has claimed that PostgreSQL outpaces competing NoSQL solutions. In that vein, Vibhor Kumar, EnterpriseDB's principal system engineer, recently produced a set of benchmarks that he believed showed PostgreSQL to be faster and less greedy with disk space than MongoDB for the same JSON dataset.

While having NoSQL features in PostgreSQL is relatively new -- PostgreSQL started adding adding NoSQL features in version 9.3, released earlier this year -- other database vendors are starting to take an interest in rolling similar functionality into their products. Oracle, too, has decided that the JSON (read: NoSQL) market is too big to ignore and recently added support for JSON data to Oracle 12. Oracle could be signaling its unease about competition from PostgreSQL, though leaving Oracle for PostgreSQL is not that easy.

Conventional SQL and NoSQL have also interpenetrated via products that place a conventional SQL querying engine in front of a NoSQL system like Hadoop. That's the pitch for Splice Machine, which claims it can provide better speeds at scale than conventional RDBMSes because of the way it exploits Hadoop's scale-out architecture. Both approaches are likely to gain ground independently, especially as signs point to SQL and NoSQL solutions growing in parallel rather than either one eclipsing the other.

[An earlier version of this story referred to PostgreSQL as a columnar, rather than a relational, database. Also, it stated that EnterpriseDB, rather than the PostgreSQL Project, had added NoSQL support to PostgreSQL, and that JSON support had been added earlier this year rather than in 2012.]

This story, "PostgreSQL ramps up its NoSQL game," was originally published at Get the first word on what the important tech news really means with the InfoWorld Tech Watch blog. For the latest developments in business technology news, follow on Twitter.

Copyright © 2014 IDG Communications, Inc.