This is the first part in a series of Blogs I intend to write on DotNetNuke Performance.
I recently spent 4 days in Redmond working in the Patterns and Practices groups Testing Lab, and I would like to express my appreciation to a number of people at Microsoft for arranging for this valuable experience, and knowledge transfer. Firstly, my thanks go to Eugenio Pace of the Patterns and Practices group for allowing us to spend 4 days in the testing lab, and to Carlos Ferre, the Patterns and Practices Performance Specialist for working with me during those four days. I would also like to thank Susan Chory and Hao Kung in the ASP.NET team for working with us to develop a Performace Plan.
The testing was carried out on a platform that simulated a Hosting Environment. We used two similar test platforms - each with a single dual processor based Web-Server and a single dual processor based Database Server. Each webserver was loaded up with 500 DotNetNuke instances - one which we used for a baseline measurement was configured with release version 4.3.5, the other webserver being configured with our current working code (4.4.0 Alpha).
I have already blogged on some of the code-bottlenecks that we had found (and fixed) using Profiler tools, and the 4.4.0 Alpha code implements most of the changes as well as some other enhancements. A discussion of the full details of these changes and the tests is beyond the scope of what can be done even in a series of Blogs.
In general, the testing confirmed what we had earlier determined - that we have made some significant improvements in both performance and scalability. Over the next few weeks we will be analysing the results obtained and developing some real guidance for developers - stayed tuned !!