Enabling real-time analytics, or shortening the time from data ingestion to action in the enterprise, is a big push in the world of databases and distributed data systems like Hadoop. Latest updates to the MapR's distribution of Hadoop are focused on real-time capabilities, specifically extending these capabilities across geographically dispersed server clusters.
Enhancements in version 4.1 include MapR-DB table replication for multiple-cluster support and for real-time disaster recovery, an API for those that code in “C,” letting them create Hadoop applications, and a new POSIX Client that boosts performance and security in real-time data applications through compression and parallel access.
Data architectures have been optimized to achieve “as-it-happens” operations through automated processes.
MapR-DB table enables multiple distributed data clusters across geographically dispersed data centers. The active-active cross-data-center capability also means its easier to deploy globally. Support for clusters replicated in multiple geographies means operational data can be stored and processed close to users or devices. All live data is immediately replicated to a central analytics cluster in real time.
“Businesses continue to push the boundaries of real-time analytics in Hadoop but can be challenged by a geographically-dispersed environment,” said Nik Rouda, senior analyst, Enterprise Strategy Group. “With the new product release from MapR, data is no longer tied to one site and can instead have global relevance. Live data updates across multiple clusters can be shared and analyzed immediately with the speed and reliability needed for enterprise operations.”
