MapR, which makes a Converged Data Platform for AI and Analytics that enable enterprises to better reap the benefits of Big Data, has announced what they are terming a breakthrough in their Data Platform technology. The enhancements extend their data fabric to cloud storage for the first time through object tiering. They also provide fast ingest erasure coding, add new security innovations, and provide a new S3 API to support next-gen applications and increase application portability.
“This platform really represents a paradigm shift for data and analytics,” said Jack Norris, SVP Data and Applications at MapR. “There’s a lot of features that provide incremental improvements, but there’s also a change in the analytics. It’s not just about a data warehouse or data lake that reports on what’s happening in the business. The analytics are now being injected into the business, so they can impact it while it’s happening. It’s about a different time frame for analytics – squeezing out the latencies and delays so you have analytics available in a whole different time frame. This paradigm will be the fuel for a tremendous amount of innovations and products.”
Norris said that the massive expansion of the scale of data makes lowering the TCO of retaining and analyzing that data more critical, and without making the model more complex.
“Having a platform with high scale and high performance is critical,” he stated. “90 per cent of the success of AI is data logistics, so the data platform is very important. This is a major data platform update.”
New data services include a native S3 Interface, for direct analytics on operational data and transparent application portability across on-premise and multi-cloud environments
“The new S3 support is part of a series of shared services under the platform,” Norris said. “The new object data service, which is S3 compatible, provides a general purpose data fabric to make organizations more agile. Related to that is a series of storage tiering features, so we can optimize for performance, capacity and cost within the platform, and do it in a way that’s transparent for the developer. They don’t have to understand how the data is being stored. The system does that.”
Fast ingest erasure coding has been added, which can now be used for capacity-optimized tiers or with high speed SSDs for an optimized analytics tier.
“We also now have audit events available as a stream, with every event tracked, so you can see how actions led to other actions, and use that data to understand trends,” Norris said. “We kept log files that did this before, but now putting them into an immutable stream provides for a more event-driven analysis.”
Another new feature is MapR Data Science Refinery.
“This is a containerized solution that makes it much easier to share data science products and outputs across an organization,” Norris stated.
The security enhancements include Secure by Default.
“Secure by Default is one-click default platform security out of the box when doing a new installation,” Norris said. “The new installer enables this, and makes it very simple, with no onerous set of tasks.”
The other major enhancement is volume-based data encryption at rest, which provides an additional means to prevent unauthorized access to sensitive data.
“These platform enhancements are a new opportunity for our integrator and reseller partners to address new approaches and architectures,” Norris said. “The cloud and containers are easy to take advantage of if the applications are very lightweight, but if they require shared data it adds to the complexity. Our data fabric now provide a common platform to support all these use cases in a way that makes all the data seem local. It also allows them to optimize security more effectively than before.”
The new release of the MapR Data Platform is entering beta testing now. It is planned to be generally available next quarter.