Business technology company IBM announced an all-flash data and file management solution for enterprises running exabyte-scale big data analytics, cognitive and AI applications.
The Hortonworks Data Platform (HDP) has certified the combined flash and storage software solution to provide more choice in selecting the right platform for their big data analytics on data processing engines like Hadoop and Spark.
IBM said the IBM All-Flash Elastic Storage Server (ESS), powered by IBM Spectrum Scale exabyte-scale software defined storage, increases data throughput by 60 percent over previous solutions.
The IBM Spectrum Scale storage software combines data from silos across the enterprise into a single view to run analytics on Hadoop and Spark. Through unifying file, object and Hadoop HDFS data storage, IBM ESS with IBM Spectrum Scale combines to shorten time to results while contributing to reducing storage capacity requirements by more than 55 percent for Hadoop and Spark workload data.
The solution provides user data with essential features such as redundant data paths, distributed erasure coding and end-to-end checksums. Based upon millions of hours of current production systems, reported over the past two years the IBM ESS design has delivered five nines of availability against unscheduled downtime.
“IBM ESS with Spectrum Scale has been developed to make data available by reducing obstacles that prevent users from quickly accessing and processing the growing oceans of differing data types stored within their disparate, isolated storage systems,” said Ed Walsh, general manager, IBM Storage and Software Defined Infrastructure.