Performance is often limited by a single component either software or hardware — the so called Performance Bottleneck. This causes the software to not be able to use the full potential of other components. In microservice architectures, a set of loosely coupled, collaborating services, this is the same. If one service has performance issues other services can be impacted. The analysis of bottlenecks is based on resource information and time-dependent metrics. The goal of this thesis is the investigation of automated detection strategies for Performance Bottlenecks in microservice architectures. The first, we defined the different kinds of bottlenecks. This information forms the basis for algorithms to find Performance Bottlenecks automatically and if possible resolve them as well. The evaluation of the detection and resolution of Performance Bottlenecks is done with the evaluation framework CASPA which allows a fairly quick setup to run a performance analysis tool on a microservice architecture. We test how good our approach is in terms of detection and resolution of Performance Bottlenecks. The evaluation shows that the approach for the types we chose were acceptable. We ended up with probability of 86 percent to detect and resolve a bottleneck in a system.