Cloudera Pitching Its Hadoop As One-Stop Shop For Enterprise Data


Cloudera was the first commercial Hadoop vendor and has raised $141 million in venture capital funding since its launch in 2008. Now, as Cloudera moves toward a possible IPO, it's pitching its Hadoop offering as a product that can handle everything enterprises need to do with their data.

Cloudera Enterprise 5, released as a public preview Tuesday and slated for availability early next year, is the basis for the vendor's "Enterprise Data Hub" vision, which looks like a major strategy shift. It hinges on the idea of customers being able to store any type of data on Cloudera, in any volume, indefinitely, Matt Brandwein, Cloudera's director of product marketing, said in an interview.

A key advantage of the Enterprise Data Hub is that customers won't have to move their data outside of Cloudera's system to do advanced analytics, data warehousing and search, and other higher-end tasks, Brandwein said.

[Related: The Top 25 Disrupters Of 2013]

"The Enterprise Data Hub is the first place data lands [in an organization] and it can scale linearly to infinity. It can store any kind of data because the file system at the core doesn't care what you put into it," said Brandwein.

Brandwein said enterprises will save money by shifting workloads from existing data systems into Cloudera's Enterprise Data Hub. Companies that must archive all their data for regulatory purposes won't need a separate storage system, and they won't have to throw away old data because it's too expensive to store, he said.

Cloudera is changing its support offerings to align with the expanded technology focus of the Enterprise Data Hub.

Cloudera Enterprise 4 includes its Hadoop distribution, its Cloudera Manager systems management software, its Cloudera Navigator data management software, and commercial support. Customers that use advanced Hadoop features like search, backup and disaster recovery, as well as Cloudera Impala, which does SQL queries on Hadoop, have to pay extra for support.

Cloudera Enterprise 5 will come with a unified support package that includes all of these features, as a way to get customers on board with the Enterprise Data Hub vision.

The Enterprise Data Hub essentially combines all of Cloudera's enterprise products into one offering. Brandwein said Cloudera isn't making it available as a single product SKU, though it is "strongly considering" doing so.

In the meantime, Cloudera plans to sell Enterprise 5 by subscription on a per-node, per-year basis, and also on a per-terabyte basis, he said.

Cloudera now has around 700 partners, and it's working with Cisco, Dell, HP, NetApp and more than 20 other vendors on Enterprise Data Hub-related integrations. Cloudera is also partnering with Informatica for advanced data warehousing.

In terms of new features, Cloudera Enterprise 5 includes In-Memory Hadoop file system caching, which speeds up data processing times and results in faster query response times. Also new are user-defined functions, which lets customers write custom code and functions to use in their SQL queries.

If they don't want to write their own code, Cloudera Enterprise 5 supports the open source MADlib library of pre-built analytics functions, Brandwein said.

Cloudera switched CEOs in June, replacing co-founder Mike Olson with former ArcSight CEO Tom Reilly, who led that company's IPO in 2007. With its Enterprise Data Hub, Cloudera is pitching its Hadoop platform as something enterprises find so useful -- and economically feasible -- that they'll end up staying a while.

PUBLISHED OCT. 29, 2013