Follow Datanami:
October 16, 2014

Datameer Introduces Smart Execution

SAN FRANCISCO, Calif., Oct. 16 — Datameer, today announced the general availability of its breakthrough technologySmart Execution and Datameer Version 5.0, at Strata + Hadoop World. The company will be demonstrating its latest release at booth number 409 in the Javits Center in New YorkOct. 16-17.

Secure Big Data Investments

Millions of dollars have been invested in developing MapReduce code for large data analytics projects. With each new technology that joins the fray, code needs to be migrated in order to take full advantage of the proliferation of tools, a time-consuming, expensive and inefficient process. Today, a number of big data computation frameworks exist to handle a variety of workloads, and that number is expected to increase as work environments grow in complexity. As the Hadoop ecosystem matures, it is becoming increasingly imperative that it can handle all data sizes and types as well as all use cases within a unified environment. The overhead of multiple environments to support a variety of use cases is complex, costly and unsustainable.

Today, Datameer, the only end-to-end big data analytics application for the Hadoop ecosystem, is eliminating this complexity by introducing Smart Execution, a technology that examines dataset characteristics, analytics tasks and available system resources to intelligently and dynamically determine and engage the most appropriate execution framework for each workload. This selection is completely transparent to the end user and does not require IT assistance or extra hardware or software.

Beta customers have seen a performance increase of up to 2200 percent in individual analytics tasks and up to a 200 percent overall execution speed improvement for their big data analytics pipeline, while consuming far less cluster resources.

“As Hadoop access is democratized, we see more Hadoop clusters used for mixed workloads. It’s no longer about big data but just about data. Making every data size and type a first-class citizen is critical to making Hadoop a valuable long-term investment,” saidStefan Groschupf, CEO of Datameer. “With Smart Execution, Hadoop truly becomes a horizontal virtualized storage and compute platform that is able to serve all analytical needs. We are excited to once again be the first to market with a completely new paradigm in big data analytics.”

First Ever Dynamic Selection

Datameer 5.0 with Smart Execution intelligently and dynamically selects the best-of-breed computation framework at each step in the big data analytics process to determine the most efficient option for the task at hand. Large data analyses will be executed in a Hadoop cluster using Apache Tez, an optimized form of MapReduce, while smaller data analyses will be executed on a single Hadoop node or using new in-memory technology. Moreover, this flexible architecture can easily incorporate new advances in the Hadoop ecosystem as they become available and are hardened to become enterprise-grade, including technologies such as Apache Spark.

“Datameer is addressing an emerging bottleneck with big data computing,” said Tony Baer, principle analyst at Ovum. “As Hadoop becomes multi-purpose, it is drawing a wider variety of workloads. Some utilize large clusters more efficiently than others. With Smart Execution, Datameer is introducing optimization that caches and directs the right workloads to the right compute framework.”

In addition to showcasing Datameer 5.0 with Smart Execution, director of product management, Matt Schumpert, will be onstage at the Strata Solutions Showcase sharing big data analytics success in the financial services industries on Thursday, Oct. 165:15 p.m. PT.

Datanami