DataStax, which offers products and services based on the open source NoSQL database Apache Cassandra, has launched DataStax Enterprise (DSE) 2.1, with increased capabilities for running a Hadoop cluster across multiple data centers, along with support for Mahout and Oracle Unbreakable Linux. The new release also adds enhancements to DataStax OpsCenter, a visual management and monitoring solution for big data platforms.
Early tests are indicating DSE 2.1 Hadoop performance is 20% faster than our previous version, according to Billy Bosworth, CEO of DataStax. “That improvement is important,” he says, “because we are seeing time and time again that the new standard for transactional applications requires an architecture capable of handling mixed workloads that can span multiple data centers and/or the cloud, without increasing operational complexity or sacrificing performance.”
The DSE platform is built on Cassandra, and is designed to manage OLTP, analytic, and enterprise search data, all in the same database cluster, according to DataStax.
DSE 2.1 now provides increased capabilities to run Hadoop operations across multiple data centers. This includes the ability to run multiple job trackers across multiple data centers and create multiple Cassandra File System (CFS) key spaces per data center, which keeps Hadoop metadata local to each data center for faster performance.
DSE 2.1 also offers full Mahout support for users who want to include machine learning and data mining capabilities in their analytic applications, as well as support for Oracle Unbreakable Linux.
In addition, according to DataStax, customers using DataStax OpsCenter will have additional administration functionality to perform various reporting functions, and manage/view the data in large Cassandra objects.
DataStax Enterprise 2.1 is currently available for download.