ECPS: An Application-Specific VM Scheduler Basing on CPU Pool Mechanism for Big Data Environment

摘要

Virtualization systems (e.g. Xen) provide powerful, flexible and scalable infrastructures for big data processing. In the virtualization system, the virtual machine (VM) scheduler is the key component to efficiently consolidate the multiple VMs on the virtual machine monitor (VMM), a.k.a. hypervisor. Xen employs the CPU pool mechanism to improve the adaptability of the VM scheduler for tasks of enormous variability and variety in the environment of big data. However, the CPU pool mechanism contains no algorithms to classify VMs or the time-slice allocation strategy, which are indispensable for a complete implementation of a high-adaptive VM scheduler. In this paper, we discuss the optimizing opportunities and introduce an application-specific VM scheduler, called ECPS (Enhanced CPU Pool Scheduler), which is a CPU-pool-based VM scheduler enhanced with VM multi-classification model and a reasonable time-slice allocation strategy. The experimental evaluation shows that ECPS can remarkably improve the performance of Xen virtualization system.