This document summarizes Apache Hadoop release 0.23, which is scheduled to be the first stable release since 0.20 in 2009. Key highlights include improvements to HDFS federation, MapReduce, and high availability. The release aims to support large clusters of thousands of machines with high concurrency. Extensive testing is being done to validate performance gains from changes like MapReduce shuffle reimplementation and optimizations for small jobs. The 0.23 branch is expected in August 2011 with an alpha release in October and production release in late Q1 2012.
2. Hello! I’m Arun… Architect & Lead, Apache Hadoop MapReduce Development Team at Hortonworks (formerly at Yahoo!) Apache Hadoop Committer and Member of PMC Full-time contributor to Apache Hadoop since early 2006 Apache HadoopRelease Manager for hadoop-0.23