Reflections on Strata: More than One Elephant in the Room

The recent O’Reilly Strata Conference held in Santa Clara was high energy, with many attendees focused on discussing details about Hadoop deployments. It has progressed a long way from the early days, when most of the conversations were about “what is Hadoop?” 

A great example of how Hadoop in production has come to the forefront was the customer panel moderated by Forrester Principal Analyst, Mike Gualtieri. He interviewed four MapR customers that represented diverse use cases across four industries: data warehouse offloading by Cisco; digital ad auctions by the Rubicon Project, which processes 90 billion ad auctions a day; sensor data and weather analysis by the Climate Corporation, which provides software and insurance products to improve farming operations; and global threat intelligence by Solutionary, as part of their managed security services. It was an excellent session; customers shared details about deployments and the successful way that Hadoop is impacting their business and customers. This session was captured on video and we expect Strata organizers to have that available in the next week or so.

There also has been a lot of discussion across the Hadoop industry about SQL-on-Hadoop options, and Strata was no exception. Our announcement with HP regarding the integration of Vertica on the MapR platform received a great deal of attention, and one that Silicon Angle specifically noted as standing out from the crowd... Why the big reaction? It’s a solution without compromises. With other approaches, the compromise could be in the SQL support; since it’s not full SQL, you’ll have to figure out what that means—but I’m sure you’ll be fine. Or the compromise could be in the data. You would have to copy the data via a connector, but don’t worry about the details of how it is duplicated, or the time lag required, or the scalability limits—you’ll be fine. Or the compromise could be in the management of a required separate cluster or the creation of dedicated separate nodes (a cluster within a cluster). Don’t worry about the extra time and complexity—I’m sure you’ll be fine.

With HP Vertica and MapR, you don’t experience any compromises. Vertica is 100% ANSI-compliant. The data is natively integrated in the MapR cluster, and finally, there are no separate clusters or nodes. This announcement was the talk of the show, as customers embraced a solution that provided answers instead of raising bigger questions.

The MapR/Vertica announcement was only one of a series of MapR announcements at Strata. We also received a lot of attention from the announcement that our latest MapR Distribution included Hadoop 2.0 with YARN, and we also received a lot of interest about our introduction of the MapR Sandbox, which makes it easy to train, develop, and test on the MapR platform.

Perhaps the best reflection about Strata is that for MapR, Strata itself reflected a lot of great work. Michele Nemschoff has written about our new logo and branding that was unvieled at the show. From the banner in the lobby, to our booth, to the balloons and other branding elements at our cocktail party, the MapR new logo and look captures the energy and excitement of MapR.

There were a lot of elephants in the room at Strata; we hope that going forward, you take the time to learn as much as you can about ours.

Streaming Data Architecture:

New Designs Using Apache Kafka and MapR Streams

 

 

 

Download for free