| By Maureen O'Gara |
Article Rating: |
|
| February 26, 2013 12:45 AM EST |
Reads: |
1,732 |
EMC and its Greenplum unit claim they have tamed some of the bears in Hadoop, the popular but difficult-to-work-with open source Big Data platform.
They’ve created a new Apache Hadoop distribution called Pivotal HD that natively integrates Greenplum’s massively parallel processing (MPP) database technology with the Apache Hadoop framework; fitted the beast out with spiffy SQL interfaces; and reportedly improved performance somewhere between 100x and 600x on a wide range of query types and workloads.
They say they’ve now got the “single most powerful Hadoop distribution in the industry” and the widgetry should be out by the end of March as a software-only or appliance-based solution.
EMC means to turn rival Cloudera into hamburger meat and then run it into the ground. It says the Hadoop pioneer can’t stand up to Pivotal’s beta; it wins every time.
The technology responsible for the performance leap is called HAWQ (say hawk), a distillation of Greenplum’s 10 years of R&D on large-scale data management put at the service of Hadoop.
Pivotal is supposed to represent the world’s first true SQL processing for Hadoop – compared, say, to existing SQL-like services on top of Hadoop – and expands the platform’s reach to SQL programmers and developers, as well as enabling compatibility with traditional BI tools.
For all the success Hadoop has had it’s been a bear to work with in part because it lacked useful interfaces and high-level tooling for business intelligence and data mining.
EMC says Pivotal HD fixes all that.
