DeFog: Adaptive Microservice Scheduling on Kubernetes Clusters in Cloud-Edge-Fog Infrastructures

DeFog is an innovative microservice placement and load-balancing approach for Kubernetes multi-cluster Cloud-Fog-Edge architectures to minimize application response times. Applications are modeled as Service Oriented Architectures (SOA) comprising multiple interconnected (micro) services. As the resources of the Edge and the Fog are limited, choosing among services to run on the Edge or the Fog is the problem this work is dealing with. DeFog focuses on dynamic (i.e., adaptive) and decentralized service placement with zero downtime, eliminating the need for coordination among the clusters. Several placement policies are tested on two realistic SOA applications to select the one that reduces application latency.