Hortonworks release HDP 1.3 with Hive enhancements
Source: Hortonworks Hortonworks has announced HDP (Hortonworks Data Platform) 1.3, the latest version of its all open source Hadoop platform. The company points out that it has achieved a steady rate of releases of the platform – 1.0 in June 2012, 1.1 September 2012, 1.2 February 2013 and now 1.3 in May – and with the latest release it has been able to focus on Hive and SQL access in Hadoop; Hive is the de facto route for accessing Hadoop data in SQL.
Hortonworks unveiled The Stinger Initiative in February as a multi-phase plan to make Hive one hundred times faster and the HDP 1.3 release sees Hive 0.11.0 included with the first phase of those improvements. The company says this should see a 50 times improvement in the performance of queries and a wider range of SQL semantics supported.
Other enhancements in HDP 1.3 include NFSv3 access to the Hadoop Distributed File System (HDFS) allowing it to be mounted as a drive on the network making it simpler to move data into the cluster, optimised Oracle and Netezza connectors, and better Sqoop support for HCatalog. Updates to HBase mean that this release also includes support for multiple masters, table snapshots and quicker recover times for better business continuity. Ambari, the monitoring and management application for Hadoop, also has improved job diagnostics and more options for customisation.
HDP 1.3 is available, packaged for Red Hat Enterprise Linux, CentOS and SUSE Linux Enterprise Server, from the download page, along with a selection of add-ons such as high availability kits, Talend's Open Studio for Big Data and Hive ODBC drivers. Release notes and documentation are also available. Source code can be found at the Apache Hadoop project.