Project Description
Windows Azure performance measurement & analytics tools.

This project is related to an artikel in the ".NET Magazine" (dutch). It is published here for the readers to download, fix or enhance the code and collaborate on the subject and the approach taken in the article.

This is a summary of what the sample application does (strike through indicates current feature status, first complete release planned for 04/10/2011):
  1. Automatically setup Windows Azure hosted services (in various regions) using the Windows Azure Management API.
  2. Upload a package file to blob storage.
  3. Add certificates to all auto-created hosted services.
  4. Deploy the package to these hosted services (and run them).
  5. Dispatch tasks to the services using queues (in this case to calculate the mathmatical constant PI).
    1. Do work in multiple regions.
    2. Do work in different timeslots (office-hours,day/night, follow-the-sun,etc.).
    3. Move task result from local storage to blob storage.
  6. Collect the results (stored in blob storage).
  7. Compare the performance data.

Here's a schematic overview of the application architecture:

The source code uses a component for which the author gave permission. Please folow the authors reference details below for more information and license details:

TPI.EXE (Pi calculation component) created and owned by Fabrice Bellard

Last edited Mar 25, 2011 at 12:18 PM by CloudCompanion, version 17