Final Thesis: Profiling and Optimizing Performance in the Cloud
Abstract: Internet users love fast websites and hate slow ones. They use, revisit and pay for sites with good performance and abandon bad performing sites forever.
High responsiveness and optimized resource utilization are therefore essential prerequisites for a website’s popularity and financial success.
In this thesis, we profiled and optimized performance in a cloud-based web application. We improved the performance of 15 backend endpoints and doubled the frontend’s overall performance score. Optimized usage of a database abstraction framework, an aggressive caching strategy, increased batch calls, and bundle size reductions had the most significant impact on performance.
We present detailed examples of performance bottlenecks in a complex web application and show how we remedied them.
Keywords: Performance Optimization, Single Page App, Google Cloud Platform, QDAcity
PDF: Master Thesis
Reference: Oscar Rosner. Profiling and Optimizing Performance in the Cloud. Master Thesis. Friedrich-Alexander-Universität Erlangen-Nürnberg: 2021.