In this short video, David Brown, Director of vFabric Applied Technology at VMware, outlines a real-life customer use case that has GemFire powering the application with real-time data. In this use case, the customer has built a system serving online ads that correlate to how in-progress sporting events are developing. For this company, having the advertising platform work closely with real-time data was an imperative for financial success.
The use case company, a popular online gaming company in the UK, created a banner advertising platform that allows bets to be placed during a game directly through their ads. These bets are called “proposition bets”, and are computer generated offers for new bets that users following a game can accept during the match. These bets include real-time odds, and change with the developments in the game. For example, a bet might be “will the current bowler hit the wicket in the next 10 bowls” – a true real-time data problem.
The company makes money by selling the ads on affiliate websites. The company’s existing system was based on Oracle, and throughput for real-time data was limited. Because of the limitations of scale,the company was losing the ability to sign up new affiliates, expand products, and was an issue withcustomer satisfaction for existing affiliates. Essentially, its lack of ability to scale was holding back the company’s ability to make money.The company upgraded their platform with GemFire to increase its capacity to process data real-time.Not only did the company achieve its goals on expanding throughput, but also as a result they were able to provide better products by expanding their programs to apply more fine-grained logic in order to select the specific banner ad for the specific user. The company was able to use temporal, sequential,regional, contractual, and other policies in order to decide which banner to deliver.
The solution underwent extensive testing because solving the real-time data problem was core to the success of the business. In this solution, the results included:
|
About the Author: David Brown started as a Smalltalk developer at IBM and has been building or architecting enterprise systems in object languages for over 20 years. Working for both GemStone and Oracle, he has focused on the issues that arise between applications and the data they depend on in high performance and distributed systems. David currently advises customers worldwide on their application of VMWare's vFabric technologies, including GemFire, tcServer, RabittMQ and Hyperic. |