The Definitive Guide to the Xen Hypervisor

Chapter 12. Virtual Machine Scheduling

One of the core features of Xen is multitasking. The hypervisor is responsible for ensuring that every running guest receives some CPU time. As with a multitasking operating system, scheduling in Xen is a tradeoff between achieving fairness for running domains and achieving good overall throughput. Some other constraints apply in Xen, due to the nature of some of its uses. One common use for a hypervisor environment is to provide virtual dedicated servers to a variety of customers. These are likely to have some form of service level agreement associated with them, and so it is necessary to ensure that no customer receives less than his allocated amount of CPU, and, for billing purposes, to track when he ...