Pre-Virtualization: Slashing the Cost of Virtualization

Despite its current popularity, para-virtualization has an enormous cost. Its diversion from the platform architecture abandons many of the benefits that come with pure virtu- alization (the faithful emulation of the platform API): sta- ble and well-defined platform interfaces, single binaries for kernel and device drivers (and thus lower testing, mainte- nance, and support cost), and vendor independence. These limitations are accepted as inevitable for significantly bet- ter performance and the ability to provide virtualization-like behavior on non-virtualizable hardware, such as x86. We argue that the above limitations are not inevitable, and present pre-virtualization, which preserves the bene- fits of full virtualization without sacrificing the performance benefits of para-virtualization. In a semi-automatic step an OS is prepared for virtualization. The required mod- ifications are orders of magnitudes smaller than for para- virtualization. A virtualization module, that is collocated with the guest OS, transforms the standard platform API into the respective hypervisor API. The guest OS is still programmed against a common architecture, and the bi- nary remains fully functional on bare hardware. The sup- port of a new hypervisor or updated interface only requires the implementation of a single interface mapping. We val- idated our approach for a variety of hypervisors, on two very different hardware platforms (x86 and Itanium), with multiple generations of Linux as guests. We found that pre- virtualization achieves essentially the same performance as para-virtualization, at a fraction of the engineering cost.