HyperGrid CTO On How Docker Is Eating Everything And What That Means For The Channel
The Container Age
Anyone who's kept an eye on the news out of Microsoft Ignite 2016 this week has most likely already heard a lot about containerization technology pioneer Docker. That's because Microsoft announced that Windows Server 2016 will include the Commercially Supported Docker Engine at no extra cost to users. In other words, Docker's technology for making the software development process easier will serve as an integral part of the newest release of the hugely popular x86 operating system.
Hyperconverged Infrastructure-as-a-Service vendor HyperGrid is bullish on containers, too. The Mountain View, Calif.-based company, formerly GridStore, got its new name after merging in July with DCHQ, a maker of technology for managing containerized enterprise applications. Last week, HyperGrid chief technology officer James Thomason gave the keynote at the Axis Business Solutions Summit, in which he predicted that "Docker eats everything" in IT over time. What follows is our Q&A with Thomason about containers and the channel at the event, which was hosted by Portsmouth, N.H.-based solution provider Axis Business Solutions.
What do you mean when you say "Docker eats everything"?
In fairness, I said that quite a while ago, back in 2014, when Docker was still sort of more in its genesis phase than it is now, where it's becoming more productized. But back then it was pretty obvious to me that Docker was the only company positioned to finally bring containers to market and create that level of disruption. So what I mean is that, from operating systems through configuration management and Infrastructure-as-a-Service, Docker has eaten everything. And I think here, late in 2016, it's already well underway.
What does that mean in practice?
So on the operating systems side, all the OSes are supporting Docker inherently. So all of the Linux vendors have moved to incorporate Docker as part of their package management – all of them have some kind of orchestration solution around [containers] already. Microsoft is investing heavily in their Docker partnership. Docker is going to be built into the next [Windows Server] version, 2016, which comes out later this year. It's built directly into Nano [Server] so you can containerize Windows apps natively on bare metal. In configuration management, a lot of these tools just aren't as useful as they used to be. You'll still need them for doing stuff around the cluster, but in terms of automating the synchronization of configurations inside of virtual machines, you don't really need to do that in a containerized world because it's neatly incorporated or encapsulated by the container build time.
When you say "eating," do you mean containers are making life harder or worse for some vendors?
What I mean is, they are creating obsolescence or disrupting the way that these tools work. Either tools are dying because you don't need them in a containerized world, or, the tools are scrambling to try to build container support, and you're not really sure what's going to happen to them. There's just a tremendous amount of disruption.
And you said private clouds are the most disrupted by this.
VMware has been the de facto way of dealing with enterprise applications inside of the environment. [With] containers, it's not clear why you need a hypervisor -- it's not clear why you need something like vSphere to manage all of this stuff. You can kind of go back to just building clusters out of bare metal and letting the container orchestrator provision your applications for you. If you're not a service provider and you don't have strict multi-tenancy requirements, then the level of virtualization that containers give you may be just good enough. I mean, it gives you resource isolation – isolation of the CPU, and storage, and network. And for, I think, 80 percent of companies, I think that's probably good enough.
As a hyper-convergence vendor, what makes Docker such an interest for you?
We really started our genesis as a hyper-convergence company, but we recently, in July, became HyperGrid. So we were GridStore, which was a hyper-converged appliance, very much in the vein of Nutanix or any of the other hyper-converged systems. But today we're HyperGrid, and the difference with us is that we're delivering a utility compute system to the customer's premises. So we're taking an AWS-like model -- so, cloud economics and everything -- we come and build an infrastructure for you. It happens to be built on hyper-converged systems, but we're not selling speeds and feeds here. And frankly as a customer, you don't care -- all you care about is that you can deploy your app, you have an SLA and that things are working right. And in the MSP industry, you know, that's typically what MSPs do. So the fact that we can, as a vendor, give them a back-to-back SLA relationship, where our technology helps them just deploy the environment for their customer, and when something goes wrong they point the finger at us, and we fix it.
So Docker is part of what folks might be using with your infrastructure?
Yeah, so we see Docker as an enabler, frankly, for us, but also a transition that companies are going to make. So we think that in the next few years, bare-metal container clusters are going to be the thing -- the norm -- more the norm than they are today. So as companies get a handle on this and start to containerize their apps, they're going to look to put those apps on bare metal and get the performance, get rid of the hypervisor overhead, stop paying the VMware tax, etc, etc.
What should the channel know about the implications of the spread of Docker?
I think the channel needs to get educated and start being aware of how to help their companies build container clusters on metal. So the new technologies are coming out from Microsoft -- from both Server 2016 and Nano -- what Microsoft is doing on that side, and what VMware is doing in order to support containers. But in Microsoft I see it more as a native container strategy, whereas in VMware it's more of a "me also." I don't think that containers are going to play quite as neatly for VMware as they do for Microsoft, in terms of fundamentally changing the way the companies manage their innovation process.
So there are opportunities for channel partners? But some might be disrupted by what Docker is doing?
So if you're in the business of just reselling VMware and building virtualization clusters, that's something that's going to start dwindling in the very near future. Because there's less and less need, in my view at least, for full-blown virtualization. Containers are probably going to be for 80 percent of folks. So they may choose to build their own tools, because they're just consuming a Docker API to do that. They may use a tool that's open source, like Kubernetes or Mesosphere, to do that. Someone like VMware may come along and buy Mesosphere or one of these companies and try to make that the next generation of what they're doing. But I don't think that over the longer horizon, five years, we're going to be just sitting back and happily selling VMware into a heterogeneous infrastructure environment anymore.
What will the model look like down the road?
I think that in the future, he who wins the server, wins the rack, wins the data center. And certainly I'm not the only one who has this view. If you look at what Cisco is doing with HyperFlex, what we're doing at HyperGrid, that is the kind of model we're taking. It's no longer about providing that Chinese menu of services that you consume from and do your own integration and deployment -- that's the way that you ran IT for 25 years. That's yesterday. The channel has to get good at delivering that utility model into the data center environment, and really providing the same value that someone like AWS provides, in a customer data center. Because as we all know, there's just a ton of stuff that can't, for various reasons, today go to public cloud.
Can you share an example?
I was talking to the CIO of a major university medical system last year, and they're doing innovative stuff -- they're taking the raw data off of the X-ray machine, the raw data off of the MRI, doing their own image processing and delivering that to a doctor's terminal down the hallway. And they have to do that within 30 seconds, so they can't take four gigs of image data, send it to Amazon, post process it and bring it back within 30 seconds -- it's never going to happen for them. So as much as that CIO would like to use something like [AWS] GovCloud, which has all of his compliance built in, he can't. So he has to build an Amazon-like infrastructure for his development teams that are doing innovative stuff. And for them it's a competitive differentiator, because they can literally see particular types of cancer, conditions, better than the hospital down the street, because they have this platform of technology that they're building inside of their system. And if you're just buying off-the-shelf stuff, from Siemens or whoever, you're not going to have those capabilities as a hospital.
So Axis Business Solutions is an example of a channel partner that understands this sort of thing?
They have a great perspective on all the technology available on the market, and that allows them to give customers honest feedback and the ability to provide a full solution, rather than just a product that fits the customer needs at that moment. And so, thematically, it's not just about speeds and feeds and what product best fits a point solution on the market, but [instead], how do you solve your customer's business problem? And their biggest problem today, I think -- the biggest challenge we have -- is innovation, as I said in my talk. IT is going from being a supporting cast member, who just used to provide IT services or apps, to a fundamental pillar of the business that everyone has to ride on top of.
What should the channel be thinking about in particular?
[All] companies, whether they like it or not, are becoming software companies. And if they're not becoming software companies, then someone else is coming along and cannibalizing what they do. So that's the core business problem I see. And so in that world, IT leaders and business leaders just don't have the time to worry about that old stuff. And all those problems are solved. We know how to build web scale systems. No one's going to build a scale-out network better than Facebook, or Google, or Amazon. And so we just need folks like Axis to bring those known patterns in-house and deliver a true utility model for customers so that they can focus on their real problems, like managing innovation, and not yesterday's problems of how to put things together.
When you're out vetting channel partners to work with, do you feel like a lot are starting to get this, or not?
I think Axis is very much in the leading edge, categorically speaking. A lot are still just in the mindset of the last 25 years of IT. And that's understandable, because that's their bread and butter, that's how they make money. And that worked for a really long time. Just going out and finding a new product -- that solved a customer's problem a little bit better than the product they were using -- was the way that we ran IT. And the way we ran consulting, and that we ran MSPs, and that we ran channel partners. But that's not good enough anymore.
What are the new opportunities for channel partners then?
You've got to be focused on the really big problems that are facing your customers, because they're going to change right out from underneath you. And they're either going to change in a way that you really don't like if you're a channel partner -- by going to a public cloud -- and the channel's relationship with public cloud has been tenuous at best, because they can provide a really limited amount of value in that situation. But with stuff that's really sticky in on-premise, that is going to be there forever, like the medical systems we were talking about, it's simply not going to be good enough to bring in a device that does more apps than the last device, because that's not really solving what's causing pain for your customers. And they're going to seek out someone who can really build a comprehensive, end-to-end solution that delivers the value that they need, which is to be able to consume IT resources and focus on building their apps.