Playing Nice Within the Hadoop Ecosystem
Although interest in all things relating to Hadoop is running high, there's a lot of contention these days behind the scenes between Cloudera, Hortonworks and MapR Technologies concerning when and where they can add value to the open-source platform.
To mitigate some of that contention, a large number of independent software vendors banded together to form ODPi, a non-profit consortium that focuses on, among other things, how to make sure the run-time environments surrounding Hadoop remain consistent across multiple distributions of an open-source platform. At the Apache Big Data conference hosted by The Linux Foundation, ODPi revealed it has become a gold sponsor of the Apache Software Foundation (ASF) that oversees the development of Hadoop alongside countless other open-source projects. ODPi is housed within the Linux Foundation.
John Mertic, director of ODPi, said a group made up of ISVs in the ODPi consortium (which counts IBM, GE, Splunk, EMC, SAS Institute, Infosys and Capgemini among its members) is trying to make sure there are enough checks and balances to prevent Hadoop from fragmenting into a set of incompatible distributions. While each distribution uses the same core base, it's become apparent that the leading providers of these distributions are creating extensions that could ultimately force ISVs to incur higher costs by having to support applications optimized for a particular instance of Hadoop.
Aiming to avoid that outcome, the ODPi provides a way for ISVs to exercise collectively their influence on the future direction of Hadoop, said Mertic. Becoming a gold sponsor of the ASF means that ODPi is now contributing dollars to the development of projects in a way that should increase its influence, he added.
Solution providers across the channel have a similar vested interest. Each extension to Hadoop might add value, but at the same time, it increases support costs. Both ISVs and solution providers need to make certain those extensions are not being added simply to lock customers into one distribution versus another.
Admittedly, not everyone in the Hadoop community understands or even appreciates what ODPi is trying to accomplish. But as Hadoop increasingly becomes the data center of the IT universe, there's a lot riding on control over its future direction. Rather than leaving that solely to the discretion of the distributors of Hadoop, the ODPi wants to make sure that the concerns of the larger Hadoop ecosystem are always going to be considered.
Michael Vizard has been covering IT issues in the enterprise for more than 25 years as an editor and columnist for publications such as InfoWorld, eWEEK, Baseline, CRN, ComputerWorld and Digital Review.