A service mesh is a dedicated infrastructure layer that facilitates secure, reliable communication between services in a distributed system. HAFAS is a distributed off-the-shelf timetable information system maintained by Hacon, a Siemens Mobility subsidiary. Introducing a service mesh to HAFAS is a base scenario for a call of research to understand better the requirements and impacts of a service mesh on distributed systems.

This thesis examines introducing a service mesh to a distributed software system and investigates the impact on the example of HAFAS. Scenarios for HAFAS that profit from the features of a service mesh were identified and mapped to functional requirements. The service mesh technological landscape was investigated by defining additional non-functional requirements and identifying a project satisfying the requirements. A subset of prioritized scenarios emerged after investigating the previous scenarios more closely in the context of the identified service mesh project.

By introducing the identified service mesh project and the scenario subset to a HAFAS environment as a case study, valuable data on the feasibility of the operation has arisen. The resulting data allowed the investigation of value added by the service mesh and impacts on latency and resource consumption. Data analysis showed that introducing a service mesh has an acceptable impact on latency, memory, and CPU resources. The benefits gained through mTLS encryption improve the system’s overall security by introducing the security properties of confidentiality, integrity, and authenticity to HAFAS’s data in transit.

The results of this thesis help organizations that plan on adopting a service mesh to their technology stack.