Network Aware Load-Balancing via Parallel VM Migration for Data Centers

Abstract

It becomes a challenge to design an efficient load balancing method via live virtual machine (VM) migration without degrading application performance. Two major performance impacts on hosted applications that run on a VM are the system load balancing degree and the total time till a balanced state is reached. Existing load balancing methods usually ignore the VM migration time overhead. In contrast to sequential migration-based load balancing, this paper proposes using a network-topology aware parallel migration to speed up the load balancing process in a data center. We transform the VM migration-based multi-resource load-balancing problem into a minimum weighted matching problem over a weighted bipartite graph. By obtaining the minimum weighted matching pairs through the Hungarian method, we parallel migrate multiple VMs from overloaded hosts to underutilized hosts to reduce the time it takes to reach a load balanced state. The experimental results show that our algorithm not only obtains a compatible multi-resource load balancing performance but also improves the balanced time which results in at most a 10% throughput gain by assuming a large batch application running on all VMs.