Azure Unblogged - Azure Stack HCI and the Future
Apr 8, 2020 05:15 · 2458 words · 12 minute read
[MUSIC] >> Hey, I’m here with Carmen Crincoli. We talked a lot before. >> Many times. >> I know you from the Windows Server Software-Defined Datacenter program? >> Yeah, it was the program, it was the Windows Server Software-Defined solutions program. >> Okay, and then I knew you were working on Azure Stack HCI, which both programs and both solutions make me super excited. So that’s why I have you here as well. >> I mean, the Azure Stack HCI program was an evolution from WSSD. >> Okay. >> WSSD. We launched it when we launched Windows Server 2016 and the idea was just we needed a way to get customers solutions that would work with our software-defined datacenter software technologies that were just built into Windows Server, and customers to try build their own hardware that was lying around, I had really bad experiences.
01:05 - Because it turns out building high-performance, resilient storage and networking requires specialized equipment. So rather than just put the software out there and hope customers get it right, we decided to work with the hardware ecosystem, have them build solutions based on best practices and then put them together and market them as a complete ended solution. WSSD was really started that process and when time came to release Windows Server 2019, we decided to go all in on the investment and we relaunched it as the Azure Stack HCI program. We built a whole road-map and planned out to make Azure Stack HCI part of our entire edge computing plan going forward. >> Yes. >> So I’m sure a lot of people have heard of Azure Stack, which is now Azure Stack Hub.
02:01 - We have a whole category in there and Azure Stack HCI is one of those pieces and that’s going to be the direction we take this hardware solution program going forward. >> So I absolutely remember. I remember the time where we were building solutions based on Windows Server and Hyper-V and again, the customers sometimes they just ordered some random hardware. So I went through that pain we’ve like, driver’s validation, doing the right setups and everything and that’s exactly where I think the Windows Server Software-Defined datacenter program comes in. Then now Azure Stack HCI. >> That’s right. Again, just remove the amount of work a customer has to do before they can even get started. Take care of that. Then, we started it as a solution program working with OEMs to define those skews.
02:59 - But it’s really a product and it’s going to become more of a product as we go forward. So right now, the program defines the hardware and then Windows Server is installed on top of it and that’s an Azure Stack HCI solution. Going forward Azure Stack HCI is really going to be a distinct offering with a lot of value added to it. >> Okay. >> So that’s the future direction we’re taking it. So that Azure Stack HCI really becomes your hybrid hyper-converged on-prem platform.
03:30 - >> Okay, , and that’s really good to hear that there is a lot of investment in that because we see a high demand from when I talk to customers out there, I really see a high demand of people are interested in a hyper-converged solutions and they are looking forward like, what are the Microsoft offerings there. A lot of what I heard from him is, “Okay, you have this Azure Stack thing?” They still believe, “Okay, there’s only one thing in Azure Stack” Because we had Azure Stack as a name. Now it’s Azure Stack Hub.” So where would you say are little bit the differences of Azure Stack Hub and Azure Stack HCI in terms of where would you use what? Also I’m also interested in what are the options? Again, as the [inaudible] gets delivered as an appliance but which limits it down to certain configurations. >> Yes. >> Is that the same with Azure Stack HCI or how does that work? >> So well, let’s start at the beginning, how do you differentiate it? It’s easily the most common question because when you change names and have similar names, people are like, “Wait a minute, there was already a thing that was called that. Why is that different? Why do I want that?” So easily the most common question, and in fact, it’s basically why I spent all of Ignite doing was answering that question because that was the first time we rolled out the whole new family and the name for the products.
04:54 - That was just what people wanted to understand it’s more clarity than just the announcement. >> Yes. >> So what we say is Azure Stack Hub, which is just called Azure Stack previously, is really about having Cloud consistency in your datacenter. >> Yes. >> Meaning if you are already down the CICD pathway and you are developing Cloud-native applications, but you have a need to run those Cloud native applications in your own datacenter, you need Azure Stack Hub because Azure Stack Hub provides the Azure Resource Manager and the APIs that you need for those Cloud-native apps to run there. So it handles that very specific business need. Azure Stack HCI is really about just providing a hybrid model for what I would call more traditional virtualized workloads.
05:49 - So if you have a virtualization firm that is just running VMs of all shapes, sizes, and descriptions and they don’t necessarily make an easy migration or fit into an IaaS world for your business Azure Stack HCI gives you the ability to have hybrid platform that takes advantage of Cloud resources without being a Cloud resource in and of itself. >> Okay? >> So it helps people who are not far enough down the Cloud path to really want the full Cloud experience running in a rack in their datacenter. >> Yeah. Which we see a lot. >> Yes. >> I mean, that’s still not something which is uncommon. Everyone would look at, most companies have a history and they’re not like fully migrated, then everything is Cloud native and everything. So I see that is a lot. But I also found it’s like there’s many different solutions out there.
06:47 - How many do we have right now in the Azure Stack HCI catalog? >> Last time I looked, I think we were up over a 190. >> Okay. >> Different solutions from 20 something different OEMs around the world. We obviously have all of the big players that you know. But we also have a lot of regional specific OEMs who have their loyal customer base because they provide a lot of services, a lot of value to their own customers. We’re happy to have that variety in the catalog, and we can go from tiny little configurations that can sit on a desk and you wouldn’t even know that there are two Xeons in those servers, and it’s a spillover cluster with 12 terabytes of capacity in it to gigantic systems that have multiple petabytes.
07:44 - The variety is part of the point, because unlike a lot of more restricted catalogs where you just have to take what’s available off the shelf, you can spec almost anything you want in it. Listen, a lot of traditional IT people love that. >> Right. >> Yeah, I love that. That’s the world I came out of. I’ve been doing this for a long time. I like having some level of involvement in the hardware and being able to spec it to my needs and seeing what the options are. >> I mean, also, their business requirements are always different. I mean for a branch office, you can’t just go out.
08:17 - Sometimes you just need something very small or even if you go even outside of offices. You probably just want to have two-note mini systems which probably still have enough performance, but you want to have a very small system. But then if you go, you take this [inaudible] to want to probably run like high-performance systems or maybe even like high capacity. So I just think that’s what I love about Azure Stack HCI. It really gives you that flexibility to whatever the customer need is. We can basically go out and address it. >> Yeah.
It was one of the high points of a big night for us 08:50 - was during our directors session, Bernardo called us. He had Cosmos Darwin come out on stage, one of PMs responsible for Storage Spaces Direct in HCI, and he talked about that flexibility and he walked out on stage with a laptop bag and everyone’s like, “Does he know that he’s carrying his laptop on stage in front of 5,000 people?” While he was talking, he was, when it allows for exciting new form factors, he opened up the laptop bag and he pulled out a full Xeon server from Lenovo. It’s called the SE 350 that they just announced for the catalog. You’re familiar with 1U pizza box servers as we call them. It’s one-quarter of one of those. That’s right. So you can take four of them and put them on a slide and put them in a place where one server would have existed previously.
09:42 - They have all kinds of cool mounting options like you can just take it VESA mount, slap it on a wall, and it is a full blown server. It can host a GPU, right? >> Yeah. >> It’s got a 100 gig networking in it. It can do 5G. It is a real server, and it’s just not something your traditional appliance model would allow for or make sense even. The variety is very much the point of making this a broad partner ecosystem firm from my perspective. >> Now again, I am super excited about we’ve already have that, I mean about the whole portfolio we offer there.
10:22 - But now you told me you’re going to work on something new which is also still to Azure Stack HCI, right? >> Yeah. >> But you have a new position or a new goal? >> Yeah. I’m not even really changing teams exactly. I’m just changing what it is I am focusing on. So rather than running the program and working with the partners, one of the things that is very important to our engineering efforts, is that we’re making customer-driven decisions. So we’re trying to start up a customer co-engineering program, where we can build relationships with customers who are doing the next thing in their own environment, and are interested in working with us on the development of that next thing. >> Okay. >> Right.
So it’s 11:15 - a very intimate program with direct engagement with our engineering. So it’s not something where we’re just going to send you e-mail updates. >> Like a survey and that’s it. >> Yeah, exactly. How happy are you? Very satisfied. It’s none of that. It’s regular syncs with our engineering teams, with our leadership. We want to understand your environment, we want your feedback about what’s going on, and we’re trying to get that customer muscle built up in our engineering teams where every time they go into make a decision or make an investment of Dev time and PM time into work, they’re like, we’re doing this because customer X said that would make a tremendous impact on their business. >> Okay. >> So we’ve had lots of different customer engagement mechanisms over the year.
12:09 - Microsoft’s been around for a long time, Windows has been around for a long time. We’re trying this different way to get that relationship up rather than have it be highly programmatic like the previous top programs. We are like Windows Insiders, which is tremendously valuable, but it’s meant to reach millions of consumers, where millions of consumers are not the audience for HCI or Kubernetes, right? >> Yeah. >> So there’s two areas I’m most interested in right now is HDI and Kubernetes on-prem and in hybrid scenarios. So if you’re watching this, I’m sure that there’s probably going to be my contact information shared on the screen somewhere when this gets published.
12:52 - Feel free to reach out to me either on Twitter or e-mail if you can find my e-mail or LinkedIn. If you think that sounds like you, please reach out. I am happy to have discussions, and if you’re working on the next-gen thing, we want to hear from you. >> Absolutely. So we will put all the links and how you’re going to reach Carmon in the description to make sure that people can actually go and engage with you. >> Perfect. >> I have one question [inaudible] what is like, okay for a customer obviously, if I join and I will have these regular calls, do I get early access to the stuff your team is doing or what is kind of like? >> The goal really is we want to co-engineer with customers. This is an NDA program. It’s not broad.
13:41 - There’s reason I am not even giving you the details about what we’re working on. We have regular briefings with us. We share early access to technology with you. We ask that you run it and give us feedback on it. The same thing with the business direction were taking things in, our technology focus, whether or not it meets your needs. That’s the kind of focus of this type of program, and then at the end of the day, we’re hoping that we make you happy enough that, one, you want to run the final product and two, that you’re willing to stand on stage with us at some point and say, “We work with Microsoft on this technology. We’re really thrilled. Here’s what we’re doing with it.
” 14:25 - We want it to be successful for both parties involved. >> Yeah. Okay. Perfect. So I want to thank you very much for this. >> Yeah. >> Super exciting to talk about this because I’m really looking forward to see what will be coming up. For all the people who are out there watching this video, if you want to have early access to the Azure Stack HCI and want to shape the future of how we do hyperconverge and all of that, reach out to Carmon. Again, you will find the links in description underneath this video. With that, I want to say thank you and see you in the next one. [MUSIC] .