Hadoop Acceptance Growing

There is much greater acceptance of closed-source community of technology providers now building value on top of the Hadoop platform. At the recent Hadoop Summit 2012, this industry metric was shared:Â There were nine providers in the Hadoop market in 2009, and now that number has surpassed 120 in 2012.

Hadoop alone cannot solve all problems for big data in the enterprise. It is being rapidly adopted, but is being integrated with other traditional database management systems, including data warehouse/BI platforms. All the major all-purpose IT providers (Hewlett-Packard, Dell, Oracle, Cisco Systems, IBM, EMC) have their own Hadoop options.

Hadoop in the cloud will take off. It is much easier to provision resources through virtualization and it’s economically more attractive and faster to get an environment up and running.

Big data technologies allow you to keep all the detailed raw data. Before these technologies existed IT managers would aggregate and roll-up data to give users access for query and analysis. Now, the data analysts can have all the data they want to uncover real business truths.

The barriers to implementing big data and Hadoop Adoption continue to be lack of skills, security and availability of the environment (data, too). Both the open source and proprietary software development communities are now addressing this.

Conventional enterprises (banks, telecoms and retailers) are now running business applications on Hadoop. However, few are willing to share all the details, and true ROI case studies are still hard to find. However, eWEEK, is developing case studies that will appear in the future.

Big data —as defined by the four data Vs: volume, velocity, variety, value—doesn’t always specifically signify Hadoop. Petabyte-scale systems with high-velocity data creation are run on low-cost commodity hardware such as NAS storage and not always using Hadoop.

Big data technologies will eliminate offline tape storage. There is no reason to put any enterprise data on offline tape, which is cheap, but is also inaccessible and highly risky. This, in fact, may be the most hotly argued point on this list. More than half of all enterprises currently use tape storage for at least one purpose, such as archives, backup or disaster recovery.

Machine learning and artificial intelligence is on the rise. However, it is still nascent in large and typically conservative enterprises such as banks, financial services, telecoms, retail and manufacturing.

The Hadoop platform, which only does offline batch processing at this time, will begin including things like free-text search technology and GUI-based visualization tools.