Reality of Big Data for 2014

MELBOURNE, AUSTRALIA: Big Data and Fast Data will graduate from the early-adopter phase in 2014. Ovum expects a significant wave of production ramp-ups in 2014 - and with it, a growing third-party vendor (and IT services) ecosystem that creates Big Data and Fast Data tools and solutions for the enterprise data warehousing and applications markets.

This is occurring as SQL and Hadoop platforms are diversifying, adopting multiple personalities, and providing overlapping functions.

According to Ovum's latest Big Data 2014 Trends to Watch report, SQL queries can now be run against Hadoop, and many SQL databases will be able to handle JSON document-centric queries.

And, as silicon-based storage - DRAM (dynamic random-access memory) and SSD/Flash - grows more abundant, the path is cleared for a new generation of Fast Data applications that will be game-changers for the enterprise applications market.

The trends to watch for Big Data in 2014 include:

* Analytic data platforms are adding multiple personalities.

* A Big Data enterprise application market is emerging.

* There is NoSQL life beyond Hadoop.

* Data tiering is predominant in realtime data platforms.

Tony Baer, principal analyst of Software - Enterprise Solutions and author of the report, says that in 2014, analytic platforms will continue to offer more diverse personalities, offering overlapping capabilities.

"The rationale is a mix of providing convenience, leveraging familiarity with SQL, and providing unified management over diverse analytic workloads. By bundling multiple engines, management of diverse analytic workloads becomes more convenient," Baer says. "The drawback is that some of these engines or approaches are quite early in development; in many cases, implementations are still proprietary."

For Big Data and Fast Data, the path for creation of a new applications market is mostly following the "visionary" model, because a legacy of custom programming for applications did not exist. An emerging breed of Hadoop-based applications is converging multiple sources of data and providing predictive analytics for customer experience solutions that could be used for cross-selling, upselling, and providing recommendations for purchase.

Baer states: "The flip side is that Ovum expects continued growth of the Fast Data applications market in 2014. The applications heavily leverage DRAM or SSD/Flash, or both to embed complex analytics and simulation capabilities into transaction applications."

Ovum predicts that there is NoSQL life beyond Hadoop. NoSQL platforms such as MongoDB and Cassandra are poised for a breakout for Fast Data Internet applications. Ovum believes more players will join the fray and some even on Hadoop.

"Look out for document-centric data stores, such as MongoDB, as they empower the next generation of web applications. JSON may fulfill the promise that XML was supposed to deliver for structuring all that ‘other' data that SQL was never able to easily accommodate," concludes Baer.