Red Hat Contributes Software to Gluster Community
Red Hat, Inc. announced today that it has contributed software to the Gluster Community that integrates the GlusterFS open software-defined storage filesystem with the Intel Distribution for Apache Hadoop software (Intel Distribution). The resulting code, the Apache Hadoop Enablement on GlusterFS plugin, delivers a big data analytics solution that easily integrates with existing IT infrastructure. The companies jointly validated the reference architecture integrating GlusterFS with the Intel Distribution.
Red Hat has contributed code to the Apache Hadoop community supporting the Hadoop Compatible File System standard and the Intel Manager for Apache Hadoop can now configure, monitor, and manage GlusterFS as a Hadoop-compatible File System (HCFS). The Hadoop Enablement solution avoids the cost and complexity associated with creating and managing another data silo for analytics. The integrated solution was built using community-driven innovation to deliver an open and interoperable solution.
Combining the performance, security, and manageability of the Intel Distribution with the HDFS API compatibility and disaster recovery capabilities of GlusterFS, the integrated solution supports a scalable, cost-effective infrastructure for big data analytics. The Intel Distribution includes security mechanisms such as query authentication, data encryption, role-based access control, and auditing. GlusterFS maintains data locality as the cluster scales, avoids NameNode bottlenecks and the single point of failure in HDFS, and has built-in disaster recovery with its geo replication feature.
Red Hat sees the evolution of analytics extending beyond Hadoop and traditional business intelligence systems into a comprehensive view of end-to-end big data analytics. Most enterprises try to manage big data from three sources, including business-, machine- and human-generated data through a work flow that includes three types of analytics systems - massively parallel processing, Hadoop clusters, and traditional business transaction processing. This broader view of big data requires a general purpose storage repository such as GlusterFS that can store a variety of data in it's native format and serve it to a variety of analytics systems through multiple protocols. An end-to-end view of all the enterprise data and all the enterprise analytics systems offers a more comprehensive way to allow for deep business insights and help drive operational intelligence.
