Performance Analysis of Preemption-Aware Scheduling in Multi-cluster Grid Environments

In multi-cluster Grids each cluster serves requests from external (Grid) users along with their own local users. The problem arises when there is not sufficient resources for local users (which have high priority) to be served urgently. This problem could be solved by preempting resources from Grid users and allocating them to the local users. However, resource preemption entails decreasing resource utilization and increasing Grid users' response time. The question is that how we can minimize the number of preemptions taking place in a resource sharing environment. In this paper, we propose a preemption-aware scheduling policy based on the queuing theory for a virtualized multi-cluster Grid where the number of preemptions is minimized. Simulation results indicate that the proposed scheduling policy significantly decreases the number of virtual machine (VM) preemptions (up to 22.5%).