Towards Performance Management of Large-Scale Microservices Applications

Gagan Somashekar

Abstract: Microservices architecture is a popular choice for developing large-scale online applications. However, managing and debugging the performance of interconnected microservices can be challenging. This thesis develops techniques for performance management in microservices architecture based on optimization theory and machine learning. The techniques focus on solving two critical problems: configuration tuning, and bottleneck detection and mitigation.