IBM Platform Computing

Integrate Hadoop, Spark and born-on–the-cloud apps
on a shared services infrastructure

Hadoop Performance Improvement - Big Data Analytics using IBM Platform Computing Solutions

With data growing at extraordinary rates and users demanding quick, effective analysis of this data, your applications need a high performance foundation. IBM Platform Computing software is designed to improve the performance of your computing infrastructure for your most demanding applications.



Data is growing at alarming rates

With more and more intelligent and interconnected devices and systems, the amount of data you are collecting is growing at alarming rates. In some industries as much as 90 percent of that data is unstructured and growing at rates as high as 50 percent per year. To keep your business competitive, to innovate and to get products and services to market fast, you need to be able to analyze that data and extract insight from it quickly and economically. For Big data analytics, current solutions do not provide the response time for the analytics jobs decreasing user productivity and delaying decision making.


Analyze data and extract insights quickly and economically

IBM Platform Computing software is designed to improve the performance of your most demanding applications with a low latency solution for heterogeneous application integration on a shared multi-tenant architecture. Platform Symphony Advanced Edition provides a high-performance distributed runtime engine for Hadoop MapReduce applications. The software is designed to deliver high resource availability and predictability, support for multiple applications and file systems, operational maturity, SLA policy control, and extremely high resource utilization for both MapReduce and non MapReduce applications.

Based on technology that is deployed at scale in some of the world’s largest production clusters, IBM Platform Application Service Controller is a generalized resource manager that provides a shared-service backbone for a broad portfolio of born-in-the-cloud software frameworks, including Apache Spark, MongoDB and Cassandra. Platform Application Service Controller can be deployed as an add-on to Platform Symphony or as a standalone product. For environments already running open-source resource managers such as YARN and Mesos, Platform Application Service Controller provides an enterprise-grade complement, supporting a broader range of frameworks, helping avoid or eliminate cluster sprawl, and allowing customers to minimize infrastructure expansion costs.

Platform Computing resource scheduling software employs optimization techniques that allow sub-millisecond response and fast provisioning for a variety of workloads. As a result, short-running jobs spend a smaller percentage of time in the provisioning and deprovisioning steps, giving a higher useful work-to-overhead ratio. Platform Symphony has a high job throughput rate that allows more than 17,000 tasks to be submitted per second.


Enhance your Big Data & analytics applications

IBM Platform Computing solutions offer a range of benefits for your big data analytics applications:

Contact IBM


Connect with Platform Computing

Success Stories

Learn More


Datasheet


Solution Briefs


Whitepapers


Webcasts

Training, Services & Support