The Hypervisor revolution was driven by VMware, a company that has re-written how IT departments build datacentres. VMware did not invent the idea of a virtual machine, they have been around since mainframe times in the 1960’s. However VMware did bring the hypervisor into the open systems era.
Virtualising Servers with hypervisor technology delivers real freedoms and advantages over datacenters built on physical servers. You can drive down hardware costs, applications can be insulated from events like an OS hanging, also protection of applications on virtualised servers tends to be that much simpler, requiring less resource to manage.
Hypervisors abstract the management of the underlying hardware and in doing so they have been key in reducing hardware lock in. This being the case, it is slightly ironic that in recent years the hypervisor itself has become the item that actually locks users in. As an example, if you try to replicate an entire application between two different hypervisors, (e.g. VMware to KVM) it is not easy to do. The reality is that you are likely to need the same hypervisor at source and secondary location. It is not so different from the hardware lock-in people have experienced with array based replication, where they had to purchase the same enterprise arrays on primary and secondary location.
There are other parallels to hardware lock-in. Different hypervisors have different functionalities, behave in different ways and have different management interfaces. If you are very familiar with VMware for example, then if you want to also use Hyper V you need staff to develop a second set of skills.
Your choice of cloud can even be locked to your choice of hypervisor. If you plan to implement a hybrid cloud strategy, it becomes much easier to achieve if your cloud provider builds their cloud based on the same hypervisor that you use.
The other reality is that companies whose revenue is based on hypervisor license sales, have no real incentive to make it easy for users to move away from there virtual machine platform and this is where hypervisor lock in can be painful. The license cost can become very expensive.
There is Light at the end of the tunnel. Hyperconverged companies such as Nutanix, provide an open approach to hypervisor choice. Their solutions allow you to run multiple hypervisors using their own consistent management platform over the top. This immediately means that Nutanix users actually have a choice of which hypervisor they want to choose for what workload and provide tools to automatically move workloads from one hypervisor to another with a few clicks of the mouse. In addition, Nutanix have gone a step further and developed their own Hypervisor called AHV, which is based on open-source KVM. They don’t force you to use it, but it is free and there as a very viable choice eliminating hypervisor license costs.
It’s not just about cost however, Nutanix takes an open approach to hypervisor choice, because ultimately it enables their customer to develop a very open approach to virtualise and cloud-enable their datacentres. Removing this lock-in actually opens up the possibility of being able to run workloads on any platform. Freedom of hypervisor, hardware platform etc. translate to freedom of choice as you build out a public/private cloud strategy. Click here for more information.