Monday, November 10, 2014

Oracle OpenWorld 2014 Summary 3 - Big Data

Before attending this year's OOW, I hadn't paid much attention on Oracle TEAM USA winning 34th America's Cup on San Francisco Bay in 2013. But I did know the story about Larry Ellison skipping OOW keynote for America's Cup sailing last year.

While we know Larry spent more than $150 million on his two sailing boats, we might not know how Big Data helped Oracle TEAM USA won the race. According to the article from Forbes, "Each of its two boats had 300 sensors recording 3,000 variables up to 10 times per second, pumping 30,000 variables per second into an Oracle Exadata Database Machine. Sailors were outfitted with bombproof electronic tablets, wrist displays and Skipper Jimmy Spithill had a heads-up display linked into the on-board wi-fi system to constantly monitor performance and loads on the boat’s components. Comparing boat against boat during training sessions, both the Oracle and Kiwi teams were able to calculate optimal performance in any conditions. All that data gave sailors a second-by-second picture of how well they were doing relative to how well they knew the boat could perform during races."

So for OOW this year, before you could enter into the exhitition halls (Moscone South and North), you would see Oracle TEAM USA's winning AC72 catamaran parked on Howard Street. I had also got chance to take pictures with the shining trophy and sign the posters by winning team sailors.

Since Larry had taken advantage of Big Data on his own team, how would he sell Big Data to his customers?

On September 30, Thomas Kurian, Oracle’s executive vice president of product development, delivered his keynote speech focus on three major trends in business and culture—big data, mobility, and the cloud. Regarding the Big Data, Kurian introduced a new set of tools called Oracle Big Data Discovery. “It’s the visual face of Hadoop,” Kurian said.

As we all know, Hadoop is a big data platform for many companies, but is also not familiar to many professionals including Oracle DBAs because it requires expertise in a programming model called MapReduce. Oracle Big Data Discovery is a single, easy to use product, built natively on Hadoop to transform raw data into business insight in minutes, without the need to learn complex products or rely only on highly skilled resources. By using Endeca (Oracle’s data discovery engine), it lets users profile, explore, and analyze Big data as well as do prediction and correlation. “With Oracle Big Data Discovery, you will be able to explore and find patterns and problems with your data,” Kurian said.

Combined with Oracle Big Data Discovery, Oracle also brought attendees attention on newly introduced Oracle Big Data SQL. Through Big Data SQL, non-relational data on Hadoop and NoSQL databases can be accessed from within an Oracle Database using standard Oracle SQL. To lots of Oracle developers and DBAs, many of whom are pretty fluent in SQL but much less so in the technologies used to access data on Hadoop – this makes Big Data very accessible. Big Data SQL also includes a unique Smart Scan service that minimizes data movement and maximizes performance. This mechanism is similar to the smart scan facility in Exadata Storage Cells.

Like Exalogic for Middleware following successful Exadata for database, Oracle brought another appliance - Big Data Appliance. This is a specialized appliance that is an 18 node Hadoop cluster in a machine with full rack configuration (including 864 TB of raw storage capacity, 288 cores, 64 GB memory per server).The InfiniBand included in Oracle Big Data Appliance plus the storage and core design of the appliance will enable enterprises to eliminate network bottlenecks for a Hadoop cluster. The appliance includes software often required in big data projects, such as Cloudera's Distribution including Apache Hadoop (CDH) and Cloudera Manager for infrastructure administration; Oracle Linux, the Oracle JDK, the Oracle NoSQL Database (Community Edition) to support advanced analytic queries; and the popular open source R statistical development tool.

Whether to build Big Data project or buy a Big Data appliance, Oracle has both available for its customers. Because Oracle TEAM USA could win America's Cup with Big Data, will Oracle also win his customers to use Oracle Big Data?

No comments: