A load feedback based resource scheduling algorithm for IaaS cloud platform

In this paper, we integrate Hadoop onto IaaS cloud platform and propose a new Dynamic Hadoop Cluster on IaaS (DHCI) architecture. In DHCI architecture, three key modules are designed, they are monitoring module, virtual machine management module and scheduling module. The load information of both physical hosts and virtual machines are collected by the monitoring module, and can be used for designing a load feedback based resource scheduling algorithm. The basic idea of the solution is to dynamically balance the load of physical hosts and virtual machines by differentiated service. Simulation results show that our solution can balance workload and improve system performance.