Expanding Remote Access At Scale Overnight
Transcript
>> During the response to COVID 19, there was a mash rush to enable workers to remote work. This put a tremendous amount of pressure is stress, on our it infrastructure environments, not just remote work. There's examples in education, where we had to scale up the collaboration tools. Where we had to meet the needs of a remote force, or remote workforce or a remote learners, that we hadn't anticipated in a short timeframe. If you look at existing public examples of digital transformation you can take a look at Netflix, that took seven years to go from a journey of having a private data center, to having the all cloud environment.
We were asked to do what could take a year or two, in just a couple of days during the pandemic. How did we answer the bill? Many of us work from home. Organizations and went from supporting a thousand end users, to 10,000 or a hundred thousand end users. If you think about the impact to people, one of the obvious people, is the network or systems administrator. It would be nice if we can just go out to AWS, Oracle cloud infrastructure or GCP Google cloud platform, and put our entire infrastructure just, in those environments and walk away.
It's not that simple, but yet there is a between solution that we deployed in the CTL visor data center. 7, V7, typical pod in the enterprise. We wanted to know what would it take, to support our existing VDI solution and the public cloud? Is that possible? Is that a solution to the problem of going from supporting a hundred VDI users to a thousand or 10,000? Obviously all the comp providers, have that capability. So we inventoried, what did we have?
We had an existing 10 gigabit connection to the public cloud via a cloud hoteler. Meaning that we could point that 10 gig connection, or parts of that 10 gig connection to any one of the cloud providers and access those resources. So connectivity, we had it covered, but operationally the public cloud hypervisors are very different than the private cloud hypervisors. How do we move our VDI stack from one cloud provider to another one? Well, this is where the argument for X86, and continue operations and a public cloud, make sense and solve real world value.
Intel, our sponsor went from 10,000 VPN users to a hundred thousand VPN users using a very similar technique. What we did was we leveraged one of the VMware cloud solutions. So whether it's VMware cloud on AWS, Oracle cloud solution, VMware solution or VMware cloud engine or GCP, we basically took our existing operations and shifted that to the public cloud, using the software stack, that we had already trained our end users and our administrators on. We got the ability to flex up to a thousand users, 10,000 users, et cetera.
com, on the homepage, there's a link to that today. However, where's the catch like it can't be that easy, more or less it was fairly simple. There was a slight learning curve to understand networking in VMware cloud on AWS, versus our networking in the private data center. But as we started to think through more advanced solutions, I take a deep breath because I think about the enormity, of what's next. When we look at this picture, there's the temptation to say, well, we have the basics for a multi-cloud solution.
We have the connectivity in between the clouds. We have the same platform in VMware cloud solutions. Shouldn't we just simply be able to stretch that across all three cloud providers by keeping our operations? Well, this was one of the more interesting outcomes of our research. What we discovered was, we could use some of the same tools on the vSphere low level, So, telemetry data coming off of the Intel processors, OS and networking tools were more or less the same from a VM to VM comparison.
What was not the same was simply provisioning, the vSphere instances. Each one had a separate API, each one had a separate stack that they implemented. For example, Oracle cloud infrastructure doesn't come with HCX, but it allows you to choose which version of vSphere you can run. AWS comes with HCX, but you can't choose which version of vSphere you could run. You provision networking, and one way in GCP and networking a different way in AWS. It stepped on the story of, people processing technologies.
I don't have the consistent processes across all public clouds, in every way that I have in my data center. 0. What I've learned in the public cloud is that, the public cloud providers take care of all of that for me. It allows my IT team to focus on the business value and supporting applications, versus dealing with the underlay of the infrastructure. Worrying about which node support the latest version of vSphere and which nodes do not. As I've gone into phase two of the CTL visor hybrid infrastructure, we're looking at telemetry in management.
We're pushing the OEMs, VMware, HPE, Dell, and the like around these questions. How do I make this experience on prem, more like the experience in a public cloud? How do I extend my management of the private data center, VMware platform, into a singular experience in the public cloud? Some of the VMware cloud cloud foundation stuff gets me there, but it's not just about the software, it's also about the hardware, and managing security and managing life cycle and managing modern application or application modernization.
That's phase two of the CTO visor hybrid infrastructure journey. If you want to find out more about that journey, you want to follow us. com. You have questions for me, there's an about page, You can give me feedback. Talk to you next CTO dose.
