Today is a big day for Basho as we announce that the Basho Data Platform is now generally available (GA). It was late last year that the concept of the Basho Data Platform started to come to fruition. Many of our customers have been building their data platforms and services and the common patterns for their data platform architectures were emerging. Based on customer feedback it was immediately apparent that the concept of a Data Platform was ambitious, appealing and just what Basho’s customers need to be successful.
Late last year many NoSQL competitors were launching multi-model service offerings but we understood that if we were to build on Basho’s success in the market, much of which has been driven by Riak’s operational simplicity, that we were going to have to go beyond multi-model and build a unifying platform for the technology stack needed to meet the data needs of high-performance applications.
We kicked off a Data Platform workstream in Engineering in January and just over 6 months later, I’m very proud to say the 1.0 version is GA. The Basho Data Platform simplifies the development and operations of distributed big data applications – enabling high availability, scalability and operational simplicity for enterprises. The platform integrates the deployment of big data application components and complementary technologies such as Apache Spark, Redis and Apache Solr.
The team emphasized the core strengths of Riak when building the Data Platform, operational ease of use, high availability and scalability. Early on they realized we could make our Spark solution simpler by building a Leader Latch Election service on a component within Riak. This allowed us to simplify our Spark offering and ensure it was not dependent on Zookeeper. Another sign of how our engineering team continues to innovate.
There are many folks who have contributed to the development of this 1.0 release and I thank them all for their hard work and diligence. The Riak workstream stepped up to the plate by enhancing Riak with a new coverage API that determines which data resides on which vnodes and enhances read capabilities. You will see this exposed for general consumption in a new release of Riak very soon.
A few folks on the Data Platform who drove the delivery of this product deserve special mention. James Gorlick, formerly of MySpace and Rovi, brought excellent Redis and Spark usability knowledge to the team. Erik Leitch, Nasa’s Jet Propulsion Lab, led the Redis implementation. Cuyler Jones, previously Architect at Flywheel and IP Commerce, lead the Data Platform Service Manager implementation. Nick Marino implemented the Leader Latch Election Service. Matt Digan, Data Platform Lead, drove the release. I know I am forgetting some folks here but it is always difficult when you work with such a talented, dedicated group of Engineers.
One of the best benefits of developing the Data Platform was the experience of looking at the service as if we were a customer – a customer who needs to build their own data platform and services. We not only found ways to make a Data Platform in tune with customer needs but we also had revelations how to make Riak itself better.
The Basho Data Platform will continue to be an important workstream in Engineering just as our investment in Riak has been a key driver to our success. Over the next few months you’ll see follow on releases with additional functionality coming out at a regular cadence. In addition we’ve identified some very innovative and powerful services that we’ll add to the Platform this year which will clearly position Basho as a leader in the NoSQL market.
Lastly, thanks go out to the whole Basho team. This success would not be possible without the discipline of product management, the feedback from Sales and the customers they manage, the creativity of marketing, the insight of our technical account managers, and the expertise and support of all those who administer Basho. I’m very excited about the future and very happy to be working with an innovative, dedicated team.