Oracle's tightly bundled hardware and software system is based on Sun server hardware and incorporates a Hadoop distribution from Cloudera, a new Oracle NoSQL database, and an open source distribution of R statistical software.
The Big Data Appliance is designed to work alongside Oracle Database 11g, its Exadata appliance and its recent Exalytics appliance for business intelligence applications. The product supports a new set of tools from Oracle called Big Data Connectors that let enterprises move data stored in Hadoop environments into Oracle Database 11g.
Oracle announced the Big Data Appliance at the Oracle OpenWorld conference last October and was expected to start shipping the products later this year. The product is designed to give enterprises a way to quickly roll out Hadoop and other NoSQL projects without having to worry about hardware, software and integration issues.
Several others vendors such as EMC Greenplum, Dell and NetApp have similar bundled Hadoop offerings in the market. The offerings from Dell and NetApp are also based on Cloudera Hadoop, while EMC's product is based on Map R, a Cloudera rival.
Oracle's appliance validates the bundled approach and will put pressure on other companies. such as IBM, Hewlett-Packard and SAP to follow suit, said James Kobielus, an analyst with Forrester Research. All of these vendors already have considerable stakes in the big data space and can be expected to roll out similar products moving ahead.
"This is good for the market. It gives customers a choice of commercially available products," for deploying Big Data projects, Kobielus said. "It also validates the fact that core start-ups in this space [such as] Cloudera and Map R are getting better entrenched," in the enterprise, he said.
The growing interest in Hadoop among enterprise technology vendors should result in more repeatable and easy to deploy products in the near-term, Kobielus said. Until recently, every Hadoop deployment in the enterprise had to be custom built by enterprises, he said.
Companies had to buy the hardware and storage, license the software and integrate it on their own. "Having an appliance for Hadoop is like having an appliance for data warehousing," he said.
When Oracle first announced its appliance last year, many had expected the product would run an Oracle distribution of Hadoop. The fact that the company decided to go with Cloudera is significant, said David Menninger, an analyst with Ventana Research.
"Cloudera is one of the market leaders with respect to Hadoop distributions, so why fight to establish another distribution?" Menninger said.
"For Oracle's rivals, I think it means more direct competition. If Oracle had created its own distribution, it would have given the competitors a potential weakness to exploit," Menninger said. "By choosing Cloudera, it eliminates the distribution as a potential issue and therefore allows Oracle to focus on the benefits of its appliance packaging."
Oracle's new product will likely focus more attention on bundled technologies for big data implementations, Menninger said. "The message Oracle is delivering to the market is that they can be the one stop shop for all your data needs, including hardware, software, big data, small data, transaction processing, applications and analytics."
The Big Data Appliance will be available in a full rack configuration of 18 Oracle Sun servers with 216 CPU cores, 864GB of main memory, 648 TB of raw disk storage, and 40GB/sec of InfinBand connectivity for connecting multiple racks together. Oracle did not release pricing information for the product.
Jaikumar Vijayan covers data security and privacy issues, financial services security and e-voting for Computerworld. Follow Jaikumar on Twitter at @jaivijayan or subscribe to Jaikumar's RSS feed . His e-mail address is firstname.lastname@example.org.