Oracle Spatial Clients Have “Big Data” Issues

At the Oracle Spatial User Conference last week in Washington, D.C., the users took center stage to discuss their recent deployments that encompassed one huge requirement: "handle big data"

What's "big data?" Safra Catz, Oracle's Co-President said this recently, "We'll let others mess around with mere terabytes of data. We have much grander plans."

Oracle clients, like the U.S. Census Bureau have huge databases. On a "productive" day, the Bureau can produce 2.3 maps per second accessing a geospatial database of 73 million edges, 22 million faces, and 160 million nodes. These data supported Census workers who had to canvas 730,000+ assignment areas for the 2010 Census. Said Atri Kulluri, Assistant Division Chief of the Geography Division, "Our success is dependant on Oracle Spatial and the topology data model.

Chris Nelson of Time Warner Cable (TWC) and their partner IMMCO have developed "MapZilla" a enterprise -wide decision support that integrates geospatial technology.  "We use Oracle everyday for all kinds of reports. And we use a team that is very good with databases," said Nelson.

Eight years in development, the MapZilla project replaced "ink and mylar."  TWC did not have a visualization tool previously. The goals were extensive and now provides network data publishing as well as asset visualization and analysis. Through multiple versions of tools the entire geospatial solution is based upon many vendor tools including MapInfo, FME, Esri and Bentley. Nelson commented that it took longer than expected but the tool is now widely used.

Aeroports De Paris, a public company since 1945, manages all of the regional airports in Paris: DeGaulle,  LeBourget, and Orly. Olivier Dubois, manager of Oracle Spatial Consulting and Resourcing Services (OSCARS) described an extensive project that required real-time vehicle tracking, event maps,  and geolocated reports that needed to be available 24/7. The real-time component for vehicle tracking was monitoring all airport vehicles and the ability utilize a linear referencing system.

Oracle Spatial User Conference 2011 Wrap Up

The Oracle Spatial User Conference is a single day event with about 150 people and was held in Washington, D.C. last week. It was the first time the meeting was held in DC and the first time it was not held in conjunction with the GITA Annual conference.

The conference continues to be an excellent venue not just for end users but for DBA's who need to know about current product functionality. The first sessions are highly technical overviews of the product and this year emphasized the relatoinship of Oracle Spatial to the Exadata Database Machine and Exalogic Elastic Cloud. The techical training sessions for MapViewer development should not be missed if you are currently developing for web services applicaitons.

Also this year, there was ever increasing emphasis on the integration of spatial applications with Oracle's business intelligence suite. Every year for the past few, I hear more about how Oracle BI is leveraging Spatial. The Oracle Fusion project, the intergration of multiple acquisitions by Oracle over the years, seems complete. Now, some of the features like thematic mapping may not be very exciting but they appear to be quite new to some in the audience who asked some pretty basic questions. While it is still an educational process for some, it's likely one that Oracle seems ready to continue given its expansive user base of Oracle products that could eventually leverage geospatial visualizations.

Oracle Exadata Database Machine the Focus of Oracle Spatial User Conference

Jim Steiner, Oracle's VP of Server Technology, opened the Oracle Spatial User Conference by talking about the focus of the event which was "extreme spatial performance and the introduction of Oracle's Exalogic Elastic Cloud and the Exadata Database Machine. Steiner said that the Exadata Database Machine was not just a preconfigured set of hardware, but engineered explicitly to run the Oracle database and handle the most demanding geospatial workloads. He discussed the fact that 15 years ago when Oracle introduced Oracle Spatial it was a "niche technology." "The vision hasn't changed but the technology has evolved and now includes grid computing, "the cloud,"  virtualization and now database machines; "Oracle has remained true to this vision even with a changing technology landscape," said Steiner.

Another emphasis was to stress how Oracle Spatial is part of the Oracle database kernel and how it is architected to exploit the processing power bandwidth and parallelism of the Exadata Database Machine. Spatial operations can be performed in up to two terabytes of database systems global area memory. Other features of the Exalogic Database Machine is:

• Eight 2-socket database servers
• 96 database CPU cores
• 768 GB database server memory
• runs Solaris
• Database grid: up to 128 Intel cores connected by 40 Gb/second InfiniBand fabric for massive parallel query processing
• Raw Disk - up to 336 TB of uncompressed storage
• Memory - up to 2 TB
• Exadata Hybrid Columnar Compression (EHCC) - query and archive modes available 10x-30x compression
• Storage servers - up to 14 storage servers (168 Intel cores) that can perform massive parallel smart scans. Smart scans offloads SQL predicate filtering to the raw data blocks. This results in much less data transferred and dramatically improved performance
• Storage flash cache - up to 5.3 TB w/ i/o resource managemen

The meeting continues through today and more reports will be forthcoming.