MapR Technologies Inc. is betting that the impending merger of big-data rivals Cloudera Inc. and Hortonworks Inc. will mire those companies in so much organizational minutia that customers will be open to making a switch.

MapR is the sole surviving independent member of the trio of well-funded startups that emerged from the early days of the big-data management software era built around Hadoop, the open-source software whose mascot is an elephant. Today the company is rolling out a new product release along with a free “assessment service” to convince customers of its competitors make the switch to MapR.

“There’s a storm happening, and customers need a clear path forward,” said Bill Peterson, vice president of industry solutions at MapR. “We see an obvious opportunity to put together a package and a program for existing customer and at least have a conversation.”

Cloudera and Hortonworks executives said when the merger was announced last month that they will provide a three-year window for customers to continue using products from either or both companies and that a “unity release” that combines the best products from both companies will be available within a few quarters.

Peterson said customers should be skeptical. “There are a lot of parts that need to go into one release for next year to get all of this right,” he said. “What kind of innovation will happen while all that is going on?” Analysts have largely been positive on the merger, saying it opens the door for a profitable business model to emerge from the margin-starved open-source software business.

MapR’s free weeklong program, which will be promoted for about the next five months, includes a migration assessment by the company’s services organization that will result in a documented data architecture depiction along with an analysis of workflows, applications and uses cases. It will identify gaps in each customer’s big data architecture and offer an implementation plan for improvement on-premises, in the cloud and at the network edge. “We’ll leave them with a playbook for what they can get from MapR and what it will cost,” Peterson said.

In conjunction with the migration assessment, MapR is also announcing the “Clarity” release of its Converged Data Platform, a unified storage layer for a multitude of big data sources. Support for artificial intelligence platforms has been enhanced with direct support for machine learning tools such as TensorFlow, Caffe and PyTorch along with Nvidia Corp.’s new Rapids graphics processing unit acceleration platform. The MapR Data Science Refinery now supports the Helium browser for Apache Zeppelin notebooks for working with visualization packages across multiple containers.

Cross-cloud synchronization and data replication has been beefed up with support for Amazon Web Services Inc.’s S3-compatible storage instances via application program interfaces. Users can coordinate data access and flows with existing Hadoop installations to forestall the need for a wholesale replacement, the company said.

Support for containers, the small virtual machines that organizations can use to build portable applications, is improved with enhanced data access for containerized stateful applications, direct data access with native Kubernetes volume drivers and the ability to use Kubernetes directly with Nvidia containers. Kubernetes is a popular orchestration manager for organizations that use a large number of containers. Stateful applications require persistent data even when containers are shut down, something that is not provided natively in popular container instances.

Uniform management, security and data protection at the edge now includes integrated support for streaming data. A common management, data security and data protection model is now a unified part of the platform.

Photo: Flickr CC