Cloud service providers tend to favor various implementations of the open source Xen hypervisor because it’s simply not cost effective for them to pay to license a commercial hypervisor at scale. It’s not clear to what degree enterprise IT organizations will want to follow suit. But The Linux Foundation that oversees development of Xen aims to increase the appeal of Xen by making available a more streamlined version that is simpler to use.
George Dunlap, a Xen Project Contributor and a senior engineer at Citrix, says version 4.10 of the Xen Hypervisor Project includes a new user interface in addition to a trusted computing base (TCB) that has been made smaller and, by extension, more secure. The expectation is that a more compact implementation of Xen will not only consume fewer system resources, but also reduce the overall attack surface exposed, says Dunlap. Those attributes should make Xen a more attractive option, for example, in Internet of Things (IoT) projects where licensing a commercial hypervisor is likely to prove cost prohibitive, adds Dunlap.
Xen Hypervisor 4.10 makes it possible for the first time to modify certain boot parameters without the need to reboot. Guest types are now selected using the type option in the configuration file, where users can select a paravirtualization, paravirtualization hardware (PVH) or hardware virtual machine (HVM) modes. Previously, Dunlap notes, many organizations avoided paravirtualization guests, which provide more flexibility for the guest operating system in terms of hardware support, because it was too challenging to implement.
Other new additions in Xen Hypervisor 4.1 include broader support for processors from Advanced Micro Devices (AMD) and the ability to run Rkt containers on the Xen hypervisor alongside Docker containers.
“We’ve added a plug-in for Rkt containers,” says Dunlap.
The battle over hypervisors is a four-way contest between Xen, Microsoft Hyper-V, VMware and the open source Kernel-based virtual machine (KVM) project advanced by Red Hat. Each hypervisor has its merits. It’s not clear yet which hypervisor will dominate the next phase of IT. But as IT becomes more distributed in the age of the cloud and IoT, it is clear that cost is about to become a much more significant factor in determining which hypervisor to deploy where and when.