Inside Oracle

Think Big

by Alison Weiss

Oracle Big Data Appliance helps IT departments manage unstructured data.

Increasingly, executives must face the challenge of big data—huge volumes of data from a variety of sources that is growing at ever-accelerating velocity. But Oracle Big Data Appliance, an engineered system built to gather, organize, and load unstructured data into Oracle Database 11g, uses previously untapped, unstructured data to help managers spot business trends, comply with regulations, reduce costs, and gain a competitive advantage.


Certain classes of untapped data have recently garnered the attention of businesses leaders, contends George Lumpkin, vice president of product management at Oracle. These include streaming digital media, social media and blog data, and industrial data collected by sensors—in addition to traditional enterprise resource planning transactions and customer relationship management data. But managers have lacked the tools to effectively analyze these huge stores of enterprise information.

“Cost-effective solutions have not been available,” says Lumpkin. “Everyone understood big data would likely yield interesting results—but it was too complex or too expensive.”

Oracle Big Data Appliance was engineered to change that dynamic. The solution efficiently captures, organizes, and analyzes big data on a scalable, cost-effective platform and then loads results into Oracle Database 11g for advanced analytics. “We are delivering the latest techniques for managing large volumes of unstructured data with Oracle Big Data Appliance and are tightly integrating this with Oracle Database 11g, Oracle Exadata, and Oracle Exalytics In-Memory Machine—so organizations can do deeper analysis and quickly get real business value from big data,” says Lumpkin.

Oracle Big Data Appliance features an open source distribution of Apache Hadoop, software for cost-effective storage and large-scale data processing. It uses Oracle NoSQL Database, Enterprise Edition, a distributed, highly scalable key-value store. Oracle’s data integrator application adapter for Hadoop simplifies data integration from Hadoop and an Oracle database with an easy-to-use interface. Oracle Loader for Hadoop uses Hadoop MapReduce processing to create optimized data sets for efficient loading into Oracle Database 11g. Once that data is loaded, customers can use Oracle R Enterprise, which integrates the open source statistical environment R with Oracle Database 11g, to conduct data mining for deep analytics.

These solutions—apart from the Hadoop distribution—will be available separately but have been engineered as an appliance for maximum performance, bridging the worlds of unstructured data and relational databases to create one data ecosystem. “The trick of big data is seeing it all in context,” says Lumpkin.

Since the big data applications market is very immature—with few applications available off the shelf—customers will have to consider writing these applications themselves. Oracle Big Data Appliance can jump-start enterprise big data initiatives by reducing system configuration time and freeing IT resources to focus on developing applications that put big data to work.

“That’s where the competitive advantage comes from,” says Lumpkin. “If you’re a bank, and you’re the first to write an application, you may gain market share or find an untapped source of customers or opportunities within your existing customer base. If you don’t do it now, your competition will.”

Photography by Shutterstock