CERN openlab II - Platform Competence Centre

Virtualization Technology Overview

There are different approaches to virtualizing hardware. The approach taken by the Bochs x86 emulator, is to translate each instruction executed in the domain of the guest OS into equivalent operations executed in the VM. Translating each instruction leads to a significant overhead compared to native execution.

The approach taken by VMware is to translate only the instructions that need to be translated. Instructions that belong to the guest OS kernel are translated and executed in a VM, while instructions that belong to user space applications in the guest OS's domain, are executed natively. Blocks of translated instructions can be reused in order to reduce the overhead of translation. This method yields better performance than pure translation. The Virtual Machine Monitor (VMM) runs alongside the host's user space applications on top of the host OS kernel.

Xen takes a novel approach by eliminating sensitive instructions directly in the guest OSs' original source code, which is called para-virtualization. These instructions are replaced with equivalent operations or emulated by replacing them with hypercalls, which call equivalent procedures in the VMM, or hypervisor. The hypervisor runs as the most privileged kernel, while guest OS kernels run less privileged on top of the hypervisor. This method yields close-to-native performance.