SLA-based Asynchronous Coordination in Metacomputing Environments

Abstract

Federated metacomputing environments allow requesters to dynamically invoke services offered by collaborating providers in the virtual service network. They evolve into metaoperating systems that form an intermediary layer between metaprograms -programs written in terms of other programs -and the virtual metacomputer composed of collaborating services. These systems require resource management to efficiently handle the assignment of providers to customer's requests and to offer high reliability and SLA guarantees. This paper presents the SLA-based asynchronous coordination algorithm developed within the recently proposed SERViceable Metacomputing Environment (SERVME) capable of matching providers based on QoS requirements and performing on-demand provisioning of services according to dynamic requester needs. The performance analysis realized on a use case of protein structure prediction shows that the new algorithm allows us to optimize the usage of resources according to cost/time priorities and may also shorten the overall execution time in comparison to state-of-the-art methods.