<<< Previous speaker next speaker >>>

Brian Oliver, Oracle

 Brian  Oliver

Brian Oliver is a Global Solutions Architect at Oracle. He works within the Oracle Coherence Engineering group and predominantly focuses on enabling financial institutions across Europe and North America to implement massively scalable and high-performance Data Grid solutions.

Over the past 10 years he has been leading the development of large-scale multi-language and multi-currency Web, E-Commerce, Sports Gaming and Financial systems making extensive use of Java technologies. He's the founder of the Coherence SIG, probably the only regularly meeting Data Grid Special Interest Group in the world, and the Coherence Incubator, a site dedicated to providing reference implementations of architectural patterns on top of Oracle Coherence.

He's been using Java since 1996. He opens an IDE every couple of days to cook up new solutions. On alternate days he's in the air traveling around Europe and North America talking about Data Grids.

Presentation: "Oracle Performance Session"

Time: Wednesday 13:00 - 14:00

Location: Rutherford Room

Abstract: Oracle Coherence is a great foundation to build high-performance, scalable applications on top of. However, it is not a silver bullet -- you cannot just plug it into your existing application and expect it to magically transform poor-performing, non-scalable application into a scalable one. In order to take full advantage of Coherence, you need to architect your application around it. In this hands-on session we will look at several different ways to ensure that your Coherence-based application performs well, ranging from low-level optimizations, such as choosing the optimal serialization format for your objects, to architectural decisions, such as where in the grid processing of the data occurs.

Presentation: "Transforming the Reconciliation Process"

Time: Wednesday 14:15 - 15:15

Location: Westminster Suite

Business Problem:

While the front office is busy attempting to execute trades faster than ever before, down stream systems, especially in the middle-office, are starting to creak from the load. One particular area of concern is the reconciliation process. Not only does it need to scale to cope with load and unexpected trading volume spikes, there is increasing pressure to support near-real auditing to aid in the timely discovery of activities like insider trading.

Current Approaches:

What was once acceptable as an over night, weekly or monthly batch process against a database to reconcile trades, positions and cash transactions, now needs to become an event driven process to support near-real-time business monitoring.


In this talk we investigate the application of Data Grids as a means to scale out, parallelize the processing matching and provide near-real-time insight into the reconciliation process for trades, positions and cash transactions.

Presentation: "Data Grid Design Patterns"

Time: Thursday 17:15 - 18:15

Location: Westminster Suite


This presentation will guide us through the technical details of a repository of common data management, distributed computing, integration, and behavioral design pattern implementations built with or based on data grids.

We will cover commonly used and requested design patterns for distributed environments, including the command pattern, the store-and-forward messaging pattern, and the WAN replication pattern.