We've updated our technical white paper to reflect Virtual Iron 4. You can download it here.

Here is an excerpt from the Virtual Iron Virtualization Services Architecture section:

- Secure: Extremely compact with no external access except through the Virtualization Manager.
- Reliable: Based on an industry-standard Linux kernel and device drivers.
- Scalable: Automatic deployment option; no software to install or manage on each server.

This approach is different than other virtualization solutions that rely on a full host operating system or console operating system.

Figure 3 provides an architecture overview of the components in Virtual Iron’s Virtualization Services. An open source hypervisor derived from the Xen™ open source project is the first software loaded when the physical server boots. It manages all hardware resources - such as registers, memory and I/O devices. The hypervisor shares resources between virtual servers to ensure that they each receive a designated allotment of processing time and isolates resources to provide each guest operating system with the same degree of protection as if it was deployed on a separate physical host.

When the hypervisor starts, it launches the service partition as a privileged domain (also known as Domain-0) that can touch all hardware and control the hypervisor. The service partition consists of the Novell SLES 10 SP1 kernel, device drivers and user-space modules that provide functionality such as remote management and accelerated I/O. It is completely hidden from user access and has no console login or persistent state. This means there are no additional components to manage or patch, resulting in greater uptime and less administrative costs. The service partition is controlled by the Virtualization Manager through an embedded management agent.