Edge-Cloud Co-Evolutionary Task Offloading for Faulty Edge Microservices

Edge microservices encounter failures that affect the stability and availability of the entire edge microservice system. Some studies are devoted to developing efficient edge-edge offloading methods that offload faulty microservices' computing tasks to other healthy microservices. However, in the face of a large-scale microservice failure, it may become necessary to temporarily solicit assistance from the edge microservices of different regions, which complicates the offloading optimization problem and poses additional challenges to its efficient resolution. Therefore, this paper proposes an edge-cloud collaborative task offloading optimization model for faulty edge microservices to minimize task offloading delay and designs an edge-cloud twostage task offloading algorithm based on the WOSCA algorithm. The example results show that compared with the cloud-based offloading method, the edge-cloud collaborative offloading effectively reduces the task offloading delay by an average of 7.94%, ensuring the reliability of the edge microservice system.