Hortonworks and Red Hat Join to Boost Hadoop File System
- By David Ramel
- June 13, 2013
Major Big Data and open source players Hortonworks and Red Hat Inc. today announced their engineering teams will collaborate to enhance Apache Hadoop storage options for enterprises.
The companies will work together "to accelerate the enablement of the broader file system ecosystem to be used with Apache Hadoop," they said in a joint announcement from the Red Hat Summit being held in Boston.
The collaboration is a three-pronged effort, with the first part focusing on enhancing the Apache Ambari project, which aims to simplify Hadoop management via a Web UI backed by RESTful APIs. The comapnies said they aim to "support the management of Hadoop-compatible file systems, such as GlusterFS."
"With this integration," the companies said, "users will be able to provision, deploy, monitor and manage alternative file systems with Ambari."
The second part of the collaboration will focus on creating generic test suites in order to validate file system compatibility between Hadoop and other file systems. Developers will be able to freely use the open source testing blueprints for compatibility verification.
Finally, the third focus area is the integration of the Hortonworks Data Platform (recently updated) with Red Hat Storage so companies can process their stored data on the Red Hat platform. "Since Red Hat Storage is POSIX-compliant, it makes it easy to connect to the enterprise applications and run Hadoop analytics on enterprise data to reduce duplication of data and save costs," the companies said. Directly running Hadoop on POSIX-compliant storage could result in 50 percent cost savings, they said. POSIX, standing for Portable Operating System Interface, is a set of IEEE standards for maintaining compatibility between OSes such as Unix and others.
While the Red Hat Summit is closing on Friday, Hortonworks and Yahoo! are co-sponsoring a Hadoop Summit North America later this month.
About the Author
David Ramel is an editor and writer for Converge360.