One of the most developing areas in information technology at present is Cloud Computing. Based on Windows Azure platform and SQL Azure database we have the possibility to host well-scalable applications and services.
The usage of this platform and infrastructure requires new approaches to observe the context and performance characteristics of our application. In the first part of the thesis the possibilities for implementing performance measurement and logging are collected, and the best practices are presented via examples. These examples are based on the experiences of the implementation of a social e-book reader application’s server side business logic layer. From developers point of view the available Azure diagnostics products are presented as well. The second part of this paper compares the performance of the various sized instances, and the impact of the size on the infrastructure services such as distributed cache service or blob, table and queue services. The overhead of the database operations in SQL Azure is shortly analyzed as well. The performance measurements were done on a real-life application, so the results can be used for other industrial applications as well.