A scalable architecture for automated monitoring of microservices

Abstract

In this paper we propose an architecture for monitoring microservices by logging key performance metrics at both system and application levels. By using advanced analytics algorithms we can then classify the events in microservice's behaviour and automate the decision processes, thus improving the overall reliability and security. We will be using Prometheus for storing short-live metrics, OpenTSDB for long term retention and RabbitMQ for passing structured messages between various IT components which orchestrate the collection of our microservices.

Keywords

• microservice
• distributed arhitecture
• monitoring
• complexity