The successful migration to microservice-based applications by large companies such as Netflix and Amazon has popularized the idea of this architectural style and it has led to an industry wide adoption starting around 2015. The problem with the microservice architecture is the fact that identifying architectural anti-patterns has become increasingly difficult. This is due to the continuous change of the architecture and the overall growth of applications in size and complexity. Current techniques for visualization and anti-pattern detection are not adequate to help developers in identifying anti-patterns. Therefore, the goal of this study is to develop a methodology for detecting anti-patterns in the microservice architecture based on execution traces. The contribution of this research is threefold. The first contribution is the approach of distributed tracing for architecture recovery. The second contribution is the application of concepts and algorithms, developed within the field of graph theory for automatic identification of architectural anti-patterns. The third contribution is the application of our approach to a large scale (400+ services) industrial application.