Virtual Iron (VI) has released version 3.6 of their product, which is cleverly named "Virtual Iron" by the way. I've been in communication with these folks for years and have always thought that the technology they're offering is both powerful and innovative.
Version 3.6 has extended the VI environment with the following features:
- VLAN support allowing virtual machines to share NIC cards. This could significantly reduce the hardware costs when typical Web or interactive applications were being supported.
- Allow virtual machines to access SANs and virtual disks simultaneously allowing organizations a great deal of flexibility
- Continue to offer organizations a low cost alternative to other virtual machine solutions.
A short, but very sincere, history lesson
At first, VI offered a very sophisticated virtual machine that could encapsulate up the stack and, here's the innovation, encapsulate hardware underneath a virtual machine solution as well. They also offered management software that could help administrators and operators deal with VMware, Xen and VI's hypervisors.
At that time, VI was not only able to create a virtual machine image that contained an entire stack of software from the operating system all the way up to the application. This matched the capabilities being offered by VMware, Connectix (now Microsoft) and the fledgling open source project "Xen". They went beyond this to being able to run these images on a hypervisor that not only supported multiprocessor-based systems (up to 16 processors if my memory serves me well) but also could aggregate the resources of up to 16 different computers connected by Infiniband and create something that was part virtual machine environment and part single system image cluster. It was an interesting concept, one that clearly was way ahead of its time.
Since that time, VI has moved away from that hypervisor to supporting the Xen environment rather than their own hypervisor.
An even shorter situational analysis
VI has continued to be a very innovative company. It has sought out areas to add value to a virtualized environment. Each innovation has been brought out without having a negative impact on performance, scalability or manageability.
Although I'm sorry that they left their own hypervior behind, I fully understand VI's reasoning.
- It is silly for a company of this size to focus most of their engineering and support budget on a hypervisor when Xen, a quite suitable alternative, is available.
- The availability of off-the-shelf systems based upon powerful multi-core processors, make the need to aggregate the power of up to 16 different machines together rather small.
- Xen has garnered a great deal of industry support, support that VI could not hope to match with its own hypervisor
- VI would be able to focus its time and attention on making it easier to administer and operate virtualized environment. Organizations can often get the biggest return on their virtual systems investment with powerful automated management tools.
It is clear that VI's biggest strength is innovative engineering. It faces challenges from many suppliers of virtual systems management software and has wisely chosen to place its focus on that challenge rather than trying to build a community on its own hypervisor technology. VI is going to be forced to work hard to get its strong message out into a world in which virtualization has become a buzz word people are attaching to just about everything.
What do you think?
Does your organization have plans to deploy management software for virtual systems? What criteria is your organization using to select this software? If you're one of the folks using VI's tools, what benefits have you received that can be directly attributed to the selection and use of this software?