Oracle is concentrating on big data at its OpenWorld 2011 conference.
CEO Larry Ellison said Oracle’s new Exalytics Business Intelligence Machine is an appliance that will offer ‘business analytics at the speed of thought’. In more practical terms, Exalytics is an in-memory database appliance that makes use of DRAM for storage and parallelism to speed Business Intelligence queries. The device consists of Oracle’s Sun Fire server with 1 Terabyte of DRAM and the Intel Xeon E7-4800 processor with a total of 40 cores. By using compression this could store 10 terabytes of data, according to Ellison. The device can be connected to an Exadata database machine using a pair of 40 Gbps Infiniband connections.
The software used by Exalytics is a TimesTen memory-based database that has been tuned to make use of the parallel cores, running alongside a version of Essbase that has been optimized to make use of the memory. Essbase is multi-dimensional database that Oracle acquired as part of its takeover of Hyperion Solutions.
According to Ellison,
"The Exalytics machine not only handles relational data, not only handles multi-dimensional data, it also handles and analyzes unstructured data at the same speed. There is no response time, everything is pretty much instantaneous."
Ellison claims that running Exalytics on top of an existing Oracle database results in analytics being available 18 times faster, and when running with an Exadata database machine, that queries are 23 times faster.
This depends, of course, on having the right data loaded into memory, which relies on Exalytics’ heuristic adaptive in-memory cache. Data is migrated into memory according to the queries being put to the system.
The conference was also the venue for Oracle to show the Oracle Big Data Appliance, a system that includes an open source distribution of Apache Hadoop, Oracle NoSQL Database, Oracle Data Integrator Application Adapter for Hadoop, Oracle Loader for Hadoop, and an open source distribution of R. As the name suggests, the Big Data Appliance is designed to help customers work with very large data sets of the sort created by web logs, social media feeds, and devices such as smart meters and sensors. Oracle NoSQL Database Enterprise Edition is a distributed key-value database, while R is the open-source statistical software.
To be informed about new articles on I Programmer, subscribe to the RSS feed, follow us on Twitter or Facebook or sign up for our weekly newsletter.