Newsletters




MarkLogic Summit in NYC Looks at Big Data Beyond the Hype


Bookmark and Share

Enterprise NoSQL vendor MarkLogic recently brought its summit series to New York. Themed as “Big Data, Beyond the Hype: Delivering Results,” the one-day conference included presentations by MarkLogic executives as well as partners and customers. In his opening keynote, CEO Gary Bloom highlighted the need for a next-generation database to address the problems and opportunities posed by big data, while also cautioning that there are “a lot of shiny objects” in the market now trying to capture people’s attention that may not deliver the necessary results.

NoSQL represents a new era for enterprise computing, said Bloom, comparing its rise with the emergence years ago of distributed computing alongside mainframe computing. Describing the progression from the "hierachical era" to the "relational era,"  Bloom said that with the mainframe, users had to plan queries in advance and coordinate with IT and, with relational databases, they still had to know the structure of data in order to craft an effective query. But there is data beyond structured data, such as documents, video and email, that doesn’t fit nicely into rows and columns, necessitating a new approach in what he termed the "unstructured era." In fact roughly 80% of data is unstructured, and more recent analyst reports actually put the percentage at 82%, Bloom added. The new generation of technology is all about a search paradigm, he said.

Bloom showcased examples of customers using MarkLogic to solve data problems such as the Centers for Medicare & Medicaid Services (CMS) dealing with the need related to the Affordable Care Act to load and subsequently access data in a range of forms for 30-50 million citizens from 50 states, the County of Fairfax in Virginia delivering on the promise of open government, the FAA working on airline safety issues with real-time monitoring and dashboards, as well as the BBC website which needed to handle the deluge of data coming in from the London Olympic games and also allow people to have a customized user experience. There are hundreds of use cases in production and development, said Bloom, who noted that the diverse scenarios include competitive intelligence, content delivery, counter terrorism, digital asset management, fraud detection,  reference data management, search and discovery and social media analysis. 

Unlike other NoSQL vendors, said Bloom, MarkLogic puts a premium on enterprise-grade features, and with MarkLogic 6, released earlier this year, a big focus was accessibility.

In comments after his keynote, Bloom emphasized that the focus on enterprise NoSQL is not new for the MarkLogic technology, but new in how the company is positioning itself. The company started focusing aggressively on the enterprise because of what he described as the “science projects” going on where people were building applications and thinking that they did not need features like government grade security and backup and recovery. “They were thinking these were just ancillary applications, and they were getting themselves into trouble because they became mission-critical.”

MarkLogic’s approach has two main areas of focus, he said. One is it is a new generation database, and two, you still need enterprise capabilities. “For us, that is a real differentiator.” Because his background is in the database world, Bloom said, “I don’t think when you go to a new generation of database, you necessarily throw away all the key features you had in a prior generation – and I characterize enterprise-grade features as key features.”

Commenting on where Hadoop fits into the picture, Bloom said, “We are doing a few things. We are distributing Hortonworks Hadoop and then we announced native support of the Hadoop distributed file system.  We continue to evaluate as a business the other projects we want to embark on around Hadoop integration. It is a huge topic in the federal government and in fact, some people, when they think of big data, think of Hadoop. Hadoop is part of the stack, but it is not a standalone solution. It is not your big data solution, but if you put a big data file system together with a big data database, you get a pretty powerful outcome.”

For more information on MarkLogic, go to www.marklogic.com.


Sponsors