Hortonworks Advances Connected Data Architecture Strategy

At Strata + Hadoop World, Hortonworks showcased its technology solutions for streaming analytics, security, governance, and Apache Spark at scale.

“We are bringing to market a very differentiated approach toward big data in a connected data platforms-oriented approach,” said Matthew Morgan, vice president, Product and Alliance Marketing at Hortonworks. Morgan explained Hortonworks’ viewpoint that as customers increasingly look to deploy cloud-based solutions, Hortonworks’ connected data architectures and security, governance, and data access capabilities will help support integration across multiple clouds and the data center spanning data rest, data in motion, and data at the edge.

“The proliferation of modern applications, whose logic is driven by data, is creating the need to do storage and compute at the point of its inception and leveraging the cloud for proximity. These modern applications require you to go to a connected architecture where you are going to have more than one lake and the lake is going to exist both in the data center and in the cloud, and you need to manage all of it from a connected perspective,” said Morgan.

According to Hortonworks, 55% of financial services companies and 75% of retail companies in the U.S. Fortune 100 are Hortonworks customers. Many of these customers are exploring cloud-based deployments as part of a connected data platforms architecture.

These customers are deploying massive new modern data applications that are optimizing their supply chain and their retail initiatives, said Morgan, to support those applications, they require a connected approach because the sensors are out at the edge, they are in the field, and the processing on that data coming off the sensors is either at the edge or in lakes in the cloud. “That is what is inspiring us and equipping us with the roadmap to deliver this type of technology.”

Hortonworks refers to the connected architecture as a “data plane,” encompassing all data sources for both data in motion as well as data at rest, said Morgan. “We think that the concept of the data plane will become as commonplace as a data lake has become.”

As part of the connected data platforms architecture approach, Hortonworks has added new capabilities in Hortonworks Data Platform (HDP) 2.5 to deliver on the security, governance, and access requirements for connected data workloads. According to the vendor, these capabilities include Dynamic Security and Governance to allow enterprises to classify and assign metadata tags, which enable various access policies to be enforced; Enhanced Streaming Analytics capabilities to enable machine learning and actionable intelligence in modern data application; Enterprise Apache Spark at Scale (Apache Spark 2.0 technical preview) to deliver agile analytic capabilities for data scientists around the globe;  and Apache Hive with LLAP (technical preview) enables as fast as sub-second SQL analytics on Hadoop to help ensure enhanced performance for large data sets.

Separately, at Strata, Hortonworks also announced updates to the Microsoft Azure HDInsight Hadoop Cloud solution. As part of its collaboration with Microsoft, Hortonworks says the updates will allow customers to achieve higher big data query speeds, provide a more secure Hadoop solution in the cloud, and allow for an easier experience for administrators to spin up third-party ISV applications. 


Subscribe to Big Data Quarterly E-Edition