Moving to the cloud! Let me be a little more precise and say moving to the public cloud. This concept has really been embraced and thrives in the consumer market, but will this concept really take off in the corporate world, and really, should it? One of the main concepts of virtualization, in the beginning, was the ability to consolidate physical systems into a virtual environment to shrink the overall footprint, as well as to be able to take advantage of and use all available compute resources in a physical server, and to have centralized control of the computer, storage, and networking resources.
That was the beginning, and quickly virtualization workloads moved from test/development into full production workloads as the acceptance of virtualization become more and more prevalent. The next phase along the way was the concept of virtualizing the desktop and user experience (VDI). One of the more prevalent VDI concepts was that of being able to maintain the control and security of our data, with the idea that if a laptop was lost, the corporate and private data would remain secure back at the datacenter, minimizing the impact from loss or theft. You used to hear story after story of a laptop with patient, account, or otherwise private information getting misplaced or stolen, and it was always a PR nightmare for the companies that found themselves in this position.
From this string of embarrassing security breaches as well as the lack of control over what data is where and how it is secured, various policies were put into place to do things like add hard drive passwords and encrypt all data on laptops. These steps were good first steps to help secure portable data, but centralized control became the law of the land to control, safeguard, and keep the private data inside the datacenter.
“Private data secure inside the datacenter” is exactly the point I want to make. We have spent the last few years consolidating and de-duping all the important data into the datacenter, and I do not see a big push to move the important data from our private areas and into the public space. There are areas that are going to be well-suited for the public cloud. Web servers are the first that come to mind, but I think moving into the future we will see in-house virtual infrastructures or private clouds expanding, with a substantial increase in multiple hypervisor infrastructures.
We are already starting to see a dramatic increase. In the “Information Week 2013 Virtualization Management Survey”, 42% of 320 business surveyed responded that they were running multiple hypervisors, compared to 36% in August 2011. On the flip side of the coin, 58% of companies surveyed in 2012 said they were only running one hypervisor, which is down from 64% in 2011. I foresee this growth continuing and accelerating, even though we are missing one of the biggest concepts of virtualization, centralized management and control, as well as defined industry standards for everyone to work with. Unfortunately, until the hypervisor wars start to wind down, I don’t think we will get too far with standardization.
Companies like Hotlink have started to fill the management void, as long as you have vCenter running in your environment, but we do not have an effective integration engine that can communicate between all the hypervisors, no matter which hypervisors you have running in your infrastructure, at least not yet. In conclusion, I think we have spent too much time and energy over the last few years bringing and keeping all the important data into the datacenter to see companies move away from keeping corporate data private. I think that virtualization will no longer be a “what hypervisor are you running” , but will have matured to “how many hypervisors are you running in your private cloud offering?”