An agent-based service adaptation approach in distributed multi-tenant service-based systems

摘要

Service adaptation aims to alleviate the runtime quality management problem of distributed service based systems (SBSs). Most of the existing service adaptation approaches are designed for single-tenant SBSs. However, modern distributed SBSs must achieve multi-tenancy due to simultaneous existence of multiple tenants. Thus, it is vital to study service adaptation in distributed multi-tenant SBSs. Currently, service adaptation has not been properly addressed in distributed multi-tenant SBSs. Existing approaches for service adaptation in multi-tenant SBSs are centralised which are not very efficient if the SBSs are large and distributed. Some decentralised service adaptation approaches, which are developed in single-tenant SBSs, may be extended to accommodate multi-tenant SBSs. These approaches, however, either incur significant communication overhead to obtain required information, or simply assume that some specific global information is already known, which is not realistic in large and distributed SBSs. To overcome the limitations of existing approaches, in this paper, a novel agent-based hybrid service adaptation approach for distributed multi-tenant SBSs is proposed, which is based on the multi-agent coalition formation technique. Our hybrid approach combines the advantages of both centralised and decentralised approaches while avoiding their disadvantages. The experimental results demonstrate the effectiveness of the proposed approach. (C) 2018 Published by Elsevier Inc.