There is a lot of buzz around the term "big data." It's a topic that Oracle is now jumping into with both feet with a new big data engineered system as well as new Hadoop and NoSQL software offerings.
During the database general session at Oracle's OpenWorld conference, Andy Mendelsohn, Senior VP of Oracle's server technologies, explained what big data is all about. Mendelsohn said big data is all about high volume, high velocity data, where most of the data is not valuable. The goal is to find the nuggets of information buried within big data that deliver some form of value to a business operation.
Due to the nature of big data, Mendelsohn noted that a traditional structured database like the Oracle database is likely not the best solution. Instead he said that NoSQL databases, which can be more scalable for unstructured data, are a better fit. To that end, Oracle is now debuting the Oracle NoSQL database to store big data.
Big data can also come from an open source Apache Hadoop system, which is now also being integrated by Oracle as part of its big data solution. The Oracle Data Integrator with Application Adapter for Hadoop generates code that helps to bring Hadoop data into Oracle's analytics solutions. Those solutions now include support for the open source R language for statistical analysis.
"Oracle Data Integrator understands how to get data out of the Hadoop file system and then move that data into an Oracle data warehouse," Mendelsohn said.
Pulling all the software together, Oracle is rolling out its new big data engineered system, which includes Apache Hadoop and Oracle NoSQL. Mendelsohn explained that the Big Data Appliance will have 216 CPU cores and up to 864 GB of memory. The storage component is similar to what is available in Oracle's Exadata database machine, delivering up to 432 TB of storage in a parallel architecture all connected with 40 Gbs Infiniband.
As is the case with Exadata, the big data system is powered by Oracle Linux.
"We have now a complete set of engineered systems for making the deployment of big data solutions," Mendelsohn said.