An optimized platform for big data and machine learning

04 August, 2017
Rick Janowski
IBM

For organization leaders seeking to deploy the most efficient platform for big data analytics, machine learning, deep learning and cognitive applications, IBM and Hortonworks have partnered to certify IBM Spectrum Scale with Hortonworks Data Platform (HDP) on x86 servers. This adds to previous certifications for IBM Spectrum Scale and HDP on IBM Power servers. IBM Spectrum Scale is the first software-defined storage solution to be certified with HDP.

Exploit big data for competitive advantage

Efficient and effective exploitation of big data assets is critical to the competitiveness of any business. This requires the right combination of analytics tools and data management.

The process starts with garnering all relevant data from wherever it resides in the organization.  This needs to be done efficiently to ensure that the data is timely and that costs are minimized. This data may come from a variety of organizational sources that may be distributed across departments and worldwide locations. In many organizations, this data is found in numerous isolated data silos in a wide variety of formats, often accessed using different protocols, further exacerbating the challenge of pulling it together into a coherent whole for analysis.

IBM and Hortonworks have worked together to create an optimized solution for big data analytics and cognitive workloads. IBM Spectrum Scale provides a data management solution that, in practical terms, has no limit on the size of the file system. This capacity is coupled with the level of file system performance required to support fast, time-critical data analytics, thanks to its parallel architecture, no metadata node bottleneck, and numerous other features designed to maximize data throughput.

IBM Spectrum Scale supports direct Hadoop system access to geographically distributed file systems without requiring that data be ingested into HDFS. This allows data to be archived and analyzed in place, saving time, ensuring that the data is fresh and reducing capital and operational costs by minimizing storage requirements. All of this is transparent to HDFS and leads to an efficient, easy-to-manage environment.

A pre-integrated solution

For ultimate ease of deployment and operational efficiency, IBM Elastic Storage Server is a pre-integrated hardware and software solution that incorporates IBM Spectrum Scale built on IBM Power Systems servers and commodity storage enclosures. It is available in both high-performance (all-flash) and high-capacity (hybrid flash plus disk) models.

IBM Elastic Storage Server includes software RAID capability that requires just 30 percent extra capacity to offer the same level of data protection that is achievable only with threefold data using other solutions. Running IBM Elastic Storage Server along with IBM Power Systems offers the most optimized hardware stack for analytics workloads.

To learn more about why Hortonworks is excited about this partnership, visit their page, here: https://hortonworks.com/partner/ibm/

The post An optimized platform for big data and machine learning appeared first on IBM Systems Blog: In the Making.