The 2011 edition of Oracle’s OpenWorld conference took place this month in San Francisco. The oil and gas/utilities track included presentations from Southwestern Energy, Santos, Transocean and Oracle’s own Melinda McDade.
Since its acquisition of Sun Microsystems last year, Oracle is now a hardware vendor. Martin Paynter of ‘Oracle-centric’ consultants Enkitec unveiled Southwestern’s deployment of the Sun/Oracle ‘Exadata’ database appliance. Exadata offers database servers, storage and network in a single enclosure, all ‘optimized for data warehouse and transactional processing. Paynter presented some compelling benchmarks comparing Exadata execution times with the ‘traditional’ approach. He was joined by Southwestern’s lead Brad Salva who described the company’s e-business (EBS) project—a migration to Oracle’s E-business R12 and Business Intelligence, and P2ES Enterprise Land, Energy Upstream and Wellcore applications. Southwestern considers the EBS to provide a single source of truth that has been easy to deploy and offers ‘extreme’ performance. The presentation contains a lot of detail on the hardware, on application and backup benchmarking and system tuning.
Santos’ Steven Benn described another Exadata deployment—as a ‘private cloud’ for production allocation across Santos’ complex network of production and refining facilities in the Cooper Basin, Australia. The real challenges here are the business risks of missing delivery deadlines, invoicing issues and getting production data ‘right.’ Benchmarking the legacy production allocation system against Exadata showed a ten times speedup—folks were impressed. Santos is now working on improving performance on its Intergraph Smart Plant engineering applications via the Exadata cloud.
Oracle’s Melinda McDade and Stewart Levin (Stanford) returned to the ‘private cloud’ metaphor with a proposal for Exadata’s use in seismic workflows. Alongside its performance, an Exadata ‘cloud’ promises virtualized resources, scalability and data parallelism. Some jobs (Monte Carlo simulations, data mining and attribute analysis) are an ‘easy fit’ to the cloud. Others (seismic imaging, reservoir and refinery simulation) are more challenging. According to the authors, Exadata can rise to the challenge—for instance with seismic data processes running on metadata in the database and trace data in dedicated parallel storage. Proof of concept migration trials on the 60 terabyte SEG Advanced Modeling (SEAM) data set are underway. Download presentations from Oracle World on www.oilit.com/links/1110_36.
© Oil IT Journal - all rights reserved.