|
09/13/2006, 2:15 PM - 3:00 PM
Speaker:
Michael Di Stefano, Vice President of Architecture, Financial Services, GemStone Systems.
One of the greatest challenges in the deployment of SOA applications onto utility computing infrastructure is dynamic and intelligent data management and provisioning. The technical focus on the first generation of SOA has been on factoring out key business functions into re-usable components and deploying them onto application servers. At the same time, many firms have embarked on strategic projects to leverage grid technology into company-wide utility computing services. While SOA will provide tremendous gains in the maintainability and business-wide availability of reusable services, early implementations have often been unsuitable for deployment onto the utility grid.
This inability of current utility grid technologies to manage the data requirements of rapid SOA deployment prevents dynamic resource allocation in response to real-time changes in business conditions. The result is that spikes in demand for a SOA application cannot be quickly serviced even as appropriate hardware in the utility grid sits idle. Quick, on-the-fly re-purposing of hardware for a wide range of applications is constrained by the lack of a pervasive distributed operational data infrastructure--one that can both intelligently interact with compute grid schedulers and provide SOA applications dynamically scalable and lightning fast read and write access from anywhere in the grid. Put simply, data constraints limit the speed at which utility grids can adapt to changing application requirements and business conditions.
GemStone's GemFire product line is designed to solve these problems and enable extremely fast and dynamic SOA service deployments onto utility computing infrastructure. The key areas we will discuss in this forum are:
* Intelligent Distributed Caching methodologies designed to make large operational and reference data sets available to large grid environments without scalability, latency, or throughput limitations.
* Real-time data distribution that is dynamic and adaptive as various SOA services are provisioned to grid nodes in response to rapidly changing business conditions.
* Strategies for smart integration of the Data Fabric with Grid Schedulers - Allowing optimizations such as data-aware routing, pre-provisioning of data in response to real-time load analysis, and intelligent node allocation/re-purposing for the fastest possible real-time SOA application scale-up.
* Distributed, pervasive, and dynamically scalable operational data transaction services available across the entire utility infrastructure, with built-in optimizations designed to adapt to usage patterns on-the-fly.
* SOA application integration into the Data Fabric through a variety of access mechanisms and API's.
Please join us as we present these key SOA on Utility Grid data solutions provided by GemFire, with a focus on our experience with the firms that are pioneering the next generation of grid technologies.
|
