In this 3-parts series of articles, David Pallmann explains how to perform grid computations on the Azure cloud computing platform. In Part 1 he presented a design pattern for using Azure for grid computing, while in Part 2 he shows how to develop such an application. In Part 3 he is going to run this application.
Read: Grid Computing on the Azure Cloud Computing Platform, Part 2: Developing a Grid Application
The author develops a grid application on Neudesic’s Grid Computing Framework which is comprised of four main components: Loader, Worker, Aggregator, Manager. Since he is using a framework which takes care of the plumbing, the author only writes code for the Loader, Worker and the Aggregator. The Loader needs some code to know how to initialize the whole grid processing. The Worker needs some code to know what to compute on each grid node, while the Aggregator needs code to know how to collect and present to final results.