TeaStore - A Micro-Service Reference Application
Simon Eismann, Jóakim von Kistowski, Johannes Grohmann, André Bauer, Norbert Schmitt, Samuel Kounev · 2019
Micro-service architectures are increasingly adopted in industry as they offer high scalability, availability and development speed. The research community has proposed a variety of approaches to automate resource scaling, placement decisions, failure recovery, parameter tuning and many other tasks enabling the self-management of micro-services. Evaluating these approaches in a realistic scenario requires a reference application that offers a range of different behaviors, as well as the necessary degrees of freedom. Existing reference applications either build on an outdated technology stack, do not implement a micro-service architecture or do not have a realistic performance behavior. Production software can usually not be used to evaluate research methods as researchers rarely have access to production software. In this demonstration we showcase the TeaStore, a micro-service reference application with a state-of-the-art technology stack. The TeaStore consists of five services with diverse performance behavior and can be used to evaluate the applicability of novel self-management approaches for micro-services and autonomic software systems in general. We give an overview of the TeaStore architecture, show different deployment options, describe how to run different load profiles and lastly how to collect monitoring data.