Key points are not available for this paper at this time.
Orchestrating microservices in Edge environments presents significant challenges due to the distributed and heterogeneous nature of the infrastructure, as well as the constraints of limited resources and variable connectivity. This paper addresses these issues by proposing a distributed framework for microservice orchestration based on a consensus algorithm. Our approach leverages a leader-follower consensus model, adapted to handle dynamic workloads and resource allocation efficiently. Through an extensive analysis of existing solutions, we identified the limitations of traditional centralized orchestration frameworks in Edge contexts, motivating the need for a decentralized methodology. The proposed framework introduces a dynamic leader election mechanism based on workload priorities and a distributed logging system for enhanced transparency and reliability. We validated our solution through experimental implementation on an Edge cluster composed of Raspberry Pi nodes, demonstrating its ability to adapt dynamically to variable workloads while ensuring consistency and fault tolerance. The results show that the framework effectively balances computational loads and meets the requirements of modern Edge computing applications.
Morabito et al. (Wed,) studied this question.