Multi-DAG Workflow Scheduling Strategy for Microservice in Clouds

초록

Microservices have been used by more and more flexible software development, and one of the most challenging problems in Microservice is workflow scheduling. It is very necessary to meet the requirements of users and effective complete workflow scheduling. The academics have done plenty of research on task scheduling problem and have proposed various algorithms in the literature. In this paper, we present a workflow scheduling strategy for multiple Directed Acyclic Graph (DAG) tasks with an objective to complete more workflows and reduce delays, which are called Multi-DAG Tasks Urgency Searching (MDTUS). And we verify that the MDTUS can reduce the cost of workflow execution by increasing completion rate and reducing delays through comparison with traditional workflow scheduling algorithms.