Follow Datanami:
January 23, 2012

Hortonworks Hocks Hadoop Upgrade

Robert Gelber

Apache Hadoop contributor Hortonworks announced Hortonworks Data Platform version 2. HDPv2 will be using the most recent version of Hadoop (0.23). According to the Apache Software Foundation, curators and cultivators of Hadoop, the newest release is enterprise ready.

The Hortonworks Data Platform, which is powered by Hadoop, is the company’s scalable open source platform for handling big enterprise and research data. As with the other Hadoop distros floating around out there, the key to the success of the platform is the ability to integrate data from just about any source imaginable and provide a more simplified way to make use of it.

The company describes how they differentiate themselves from others offering Hadoop simplification for the enterprise, noting:

“Unlike other Hadoop solutions that lock away management features within proprietary extensions, Hortonworks Data Platform includes Ambari, an open source installation and management system out of the box. Hortonworks Data Platform also includes HCatalog, a metadata management service for simplifying data sharing between Hadoop and other enterprise information systems, along with a complete set of open APIs, including WebHDFS and those for Ambari and HCatalog, to make it easier for ISVs to integrate and extend Apache Hadoop.”


On Jan.6th, when the Apache Software Foundation made news announcing Hadoop v1.0 after 6 years of development, a number of notable new features and enhancements were made. With the release of Hadoop version 0.23, improvements have been made to both HDFS and MapReduce including:

  • NextGen MapReduce (also known as YARN)
  • HDFS Federation, which allows Namenodes to act independently and without coordination with eachother
  • Splitting MapReduce JobTracker into 2 components (resource management and life-cycle management)
  • The Resource manager will now manage global assignment of compute resources for each application while ApplicationMaster will manage scheduling and coordination.

According to Eric Baldeschwieler, CEO of Hortonworks, “With more than three years of development and much anticipation, Apache Hadoop 0.23 delivers important advancements in scalability, performance, high availability and data integrit.

He continued, “Apache Hadoop 0.23 is currently being tested across hundreds of applications in the world’s largest Hadoop deployment. We are excited to make the technology advancements in Apache Hadoop 0.23 available through an easily consumable version via the Hortonworks Data Platform v2.”

HDP was created to extremely scalable and fully open-source platform for storage, processing, analysis of large scale data. Along with HDFS and MapReduce, Hortonworks Data Platform includes Pig, Hive, HBase and Zookeeper. 

Hortonworks was created by Yahoo! and Benchmark Capital to facilitate Apache Hadoop development. They provide tech support, training and certifications for vendors, enterprises, service providers and systems integrators.

Related Stories

Hadoop Hits Primetime with Production Release

RainStor Brings Database to Hadoop

Karmasphere Ushers in New Hadoop Partner

 

Datanami