MapR Creates a ‘Game Changer’: NoSQL + Hadoop
The Hadoop company is also integrating enterprise-grade search and discovery into its Big Data platform
By: Maureen O'Gara
May. 2, 2013 08:30 AM
MapR Technologies, the Hadoop house, Wednesday released M7, a Big Data platform that’s supposed to remove the usual trade-offs involved in deploying a NoSQL database on Hadoop.
The widgetry packages Hadoop with HBase, the open source NoSQL database modeled after Google’s BigTable and now used under Facebook’s Messaging Platform. It promises high availability, performance, low latency and ease of use.
MapR CEO John Schroeder (pictured) claims the fact that HBase is built-in with Hadoop is a “game changer” for people looking to leverage Big Data and support the broadest set of mission-critical applications.
The enterprise-grade NoSQL solution evidently has great scale. It can reportedly deliver over a million operations/sec on a 10-node cluster and scale to a trillion tables across thousands of nodes.
It was built from the ground up so it wouldn’t need compactions or background consistency checks to work smoothly. Because of its instant failure recovery HBase and Hadoop applications have Four9s availability.
MapR says that with M7 there are no region servers, additional processes or redundant layers between the application and the data residing in the cluster. Its zero-administration approach includes automatic region splits and self-tuning with no downtime required for any operation including schema changes.
John Webster, principal analyst at the Evaluator Group, sees enterprise-level deployments of Hadoop MapReduce rising quickly, “driven by a need to understand and potentially adopt this new business analytics platform for business applications.”
He figures what MapR has done in response is deliver “a distribution of Apache Hadoop that addresses many of the enterprise quality issues currently limiting its adoption in production data centers.”
M7 lets HBase apps access data directly “without the redundancy of extra layers of communication yielding a single, scalable and more reliable data store that offers high performance and is easier to develop to and administer.”
Sociocast, the predictive analytics company, was part of the M7 beta test and its CEO Albert Azout says, “We require a highly reliable, blazingly fast NoSQL platform to cater to our growing customer base. We have chosen M7 over other prominent solutions in the marketplace for its capability to support 24x7 operations with minimal administrative costs as well as for delivering the best performance numbers across different workloads.”
M7 supports Big Data operations ranging from batch analytics to real-time database functions
MapR’s also integrating enterprise-grade search and discovery into its Big Data platform but that widgetry won’t be available on M7 until next quarter when customers will be able to do predictive analytics, full search and discovery and advanced database operations on a single platform.
It’s using LucidWorks Search and the integration is currently in beta.
It’s supposed to work directly on Hadoop data as well as index and search standard files without any conversion or transformation. The widgetry includes snapshots and mirrors for full restores.
Integrating search capabilities into Hadoop is called an “important milestone for the industry.” Ben Woo, the managing director of Neuralytix, said, “This is an enormous step forward especially in time-sensitive processes such as fraud detection where Big Data must be searched as it streams into the enterprise.”
MapR and LucidWorks have been close since 2011 when they cut a joint marketing agreement. In February, they deepened their relationship, releasing a connector between their two products.
Reader Feedback: Page 1 of 1
Latest Cloud Developer Stories
Subscribe to the World's Most Powerful Newsletters
Subscribe to Our Rss Feeds & Get Your SYS-CON News Live!
SYS-CON Featured Whitepapers
Most Read This Week