Jonan Scheffler interviews Microsoft MVP Steve Buchanan about Microsoft’s Azure Kubernetes Service (AKS), how Microsoft's been successful at working in enterprise and open source, where he believes GitOps is eventually going to go, and his excitement in regards to AI and blockchain and how they’re going to impact the world.
Should you find a burning need to share your thoughts or rants about the show, please spray them at firstname.lastname@example.org. While you're going to all the trouble of shipping us some bytes, please consider taking a moment to let us know what you'd like to hear on the show in the future. Despite the all-caps flaming you will receive in response, please know that we are sincerely interested in your feedback; we aim to appease. Follow us on the Twitters: @ObservyMcObserv.
Open is Not Optional: Shifting Sands and Faxing on Blockchains with Steve Buchanan
Jonan Scheffler: Hello and welcome back to Observy McObservface, proudly brought to you by New Relic's Developer Relations team, The Relicans. Observy is about observability in something a bit more than the traditional sense. It's often about technology and tools that we use to gain visibility into our systems. But it is also about people because, fundamentally, software is about people. You can think of Observy as something of an observability variety show where we will apply systems thinking and think critically about challenges across our entire industry, and we very much look forward to having you join us. You can find the show notes for this episode along with all of The Relicans podcasts on developer.newrelic.com/podcasts. We're so pleased to have you here this week. Enjoy the show.
Steve Buchanan: Good. How's it going? Thanks for having me on.
Jonan: I'm so glad that you could join us. It's going better and better, I guess. I'm feeling optimistic for the world in many respects, not so much in others but on average, I'll take it. It's been a rough couple of years.
Steve: I agree.
Jonan: And every once in a while, the sunshine breaks through the clouds these days, which it had not been doing for a long time.
Jonan: So yeah, I'll take it. So, Steve, this is your first time on the show. And in that regard, if you could provide a little background maybe for our guests, who are you? How did you end up here?
Steve: Yeah. Awesome. Well, first of all, I'm an Azure MVP. I've been an MVP with Microsoft for nine years. I'm also in the consulting space. And so I'm one of those rare directors who...I still do some hands-on: I lead teams, I do solutioning, I do a lot of stuff. And I'm a Director in the Azure space as well as Container. So I do a lot of work with AKS, GitOps doing work with that, Azure DevOps, just a ton of different stuff. I like what I do. It has a lot of variety in the cloud.
Jonan: I imagine being in that type of role, you get to play with a lot of different technologies. What is AKS, by the way?
Steve: Azure Kubernetes Service. So it's Microsoft’s managed Kubernetes service.
Jonan: So if I want to go and spin up a Kubernetes cluster with a button, I can find that on Azure; it's called AKS.
Steve: Yes. Or you could use one line of Terraform code, spin it up that way. [laughs]
Jonan: Yes. I've heard of this Terraform thing. That's pretty good. So you've been in Azure for a long time. You've got to watch a transformational period for Microsoft. When I came out of high school, Microsoft was THE employer. Everyone who was programming…we were all sitting around hacking on our TI-85 calculators and writing Solitaire and Blackjack games, and we wanted to go work for Microsoft. And then suddenly, we went into the open sources of virus era, [laughter] and it became less popular with developers for a long time. But they've turned that around. I would not have expected 15 years ago that I would feel the way I feel about Microsoft and their efforts in software today. I think that they've effectively pulled a 180 as far as embracing open source technology and the developer community. It's been pretty impressive. And you've been involved with the ecosystem for like a decade now. What do you think made that happen? How would you even do that thing? I don't know. But they've managed it.
Steve: They changed out of survival, right?
Steve: Like you just said, there was a shift where for a long time it was like, “Oh, Microsoft? Stay away from them, from their products, working there,” all of it. So they had to shift. They had to make a shift. They saw this thing called cloud after AWS hit the market and was being successful. They saw that. They saw the movement of DevOps. They saw a lot of things coming as well as what was happening with open source and that this is maybe a stretch, but taking over the world. So it was interesting. Microsoft had to shift culturally inside. And so I've been out to the campus, and I've had some conversations with some of their program managers and stuff, and they had to shift the way they exist internally. I think a lot of people don't know that or don't really know the inside story, and I think they've shared some things publicly. But if you think about it, they went from a massive enterprise organization that shipped software every three years to we have to move to this true DevOps model and not just sell it, but we have to live it. And by the way, we're shifting to DevOps, and we're shifting to cloud because we want to be successful in cloud, so we have to deliver. We have to ship all the time, not just we can ship this year, and we got three more years to work on the next version of the product. [chuckles] So it was a totally different culture shift. So they went after cloud. They had to change the way they did things internally to more DevOps.
And then also, they started to open up to open source. And now, at this point in time, I think they're the largest contributor to Linux from a code perspective. And also, if you look at Azure Cloud, over 60%, and maybe the number is even higher now, of workloads running on Azure are Linux-based. And then you throw Kubernetes in there, and that's like the fastest growing service on Azure since they released it. It's just crazy. And you look at the people that are working at Microsoft now, there are some really big names in that open source space. And so for me personally, being in the Microsoft ecosystem for so long…since I came out of school, most of my jobs were Microsoft-focused. But even in college, I had an interest in open source. So in college, I got Linux+ certified. I was always dabbling with Linux and various random open source projects, products, or whatnot, but all my jobs were primarily Microsoft-focused. So for me, it was super exciting to see this change because I've always been a fan of the stuff happening in the open-source community, but it's always been like, okay, you tinker with that stuff, that's cool. And maybe you do it here and there for your job, but you're primarily doing Microsoft stuff, but now it's like the best of both worlds. I can't even remember the last time I honestly touched a Windows Server, but I'm working in Azure every day. If I try to think back, it's probably been like a year or two. It blows your mind.
Jonan: I can't imagine what the Microsoft of 2000 would think about the world today where I’m like, someday you're mostly going to run Linux. Like, 60% of your work is going to be done in a Linux environment.
Steve: But if you think about it, if Microsoft didn't make all those shifts, shift to focus on cloud, a shift in their internal and the way they do things to more of a DevOps model, and a shift to be an open-source, would Microsoft be doing as well as they are today?
Jonan: No way.
Jonan: There's no doubt in my mind the answer is absolutely not. And Azure is actually growing fast. I still quote regularly that AWS runs half the internet, but that's no longer the case. I think the last time I saw that -- It was the case that 50 some odd percent of all internet services were running on AWS, which is why U.S. East going down gave everyone the day off. It was like, well, the internet's broken in a way none of us can fix. Huge ups to AWS; you all have a good couple of days fixing that.
Jonan: Today, you are looking at a situation where I think it's 30%+ for AWS and 15% for Microsoft, if not more, and rapidly growing. Azure has really done a good job of expanding. And I think that's largely in part to services like AKS and being able to run Linux in a Windows development environment suddenly, right?
Steve: It is. Microsoft's been really good at working with the enterprise and in that enterprise space. So a lot of enterprises finally started jumping on the cloud train. And now that they're there, Azure is more friendly, I would argue, to enterprise. And then also what we've seen in enterprises is they're adopting more and more open source technologies as these open source technologies start to mainstream, things like Kubernetes and things like Terraform. My goodness, three years ago, a lot of enterprises were like, “Oh, Terraform, that sounds cool.” Now today, when you go in to work with an enterprise, they’re like, “Our standard is Terraform. I hope you have Terraform skills because that's what we use here.” So Microsoft with Azure being friendly to that and spinning up services to support those things and also being friendly to developers with things like VS Code. My goodness, I've seen VS Code just explode. And whether you're doing Python or Go or you're doing some .NET stuff, whatever you're doing, VS Code works well with it. It's amazing.
Jonan: Yeah. VS Code has been a huge boon to the learners’ community, and I am forever grateful to Microsoft for having produced that project. I think the Windows WSL?
Steve: WPF or?
Jonan: Windows Subsystem for Linux. That's what it was.
Jonan: I kept trying to make this into WLS. The Windows Subsystem for Linux has really changed things. I've recently -- We do a lot of streaming on this team, and the streaming community largely started on Windows because everyone in the gaming community is using Windows. So a lot of the software and the hardware even only works with Windows, and they won't even tell you. You would order a piece of hardware, and they're like, “Oh yeah, it works great on Windows, impossible to use on a Mac. Sorry, we didn't mention that in the product description or anywhere else.” [laughter] It's been a long time since I've thought to check. You used to go and buy a mouse, and you had to look at the box to see if it was a Mac mouse or a Windows mouse. But you don't care anymore. But being able to work in Windows and recreate my workflow from working on a Mac for a really long time -- I've been writing software professionally for about 11 years now. And that entire time, I worked on a Mac because I wanted that Linux experience, and being able to then take that over to Windows has been awesome. Actually, I've been really impressed. I'm still disappointed, Microsoft, if you're listening, in your emojis.
Jonan: I want to choose my emojis because I need some options. I'm not digging the thick, black lines on the emojis, but beyond that, I have few complaints. It's been pretty impressive to see that transition. We were talking a little bit about this, about how we no longer have to choose when we're looking at open source technologies or Microsoft technologies or any other large enterprise. The piece I think that was missing for the enterprise community was that open source is scary to them, right?
Steve: It can be.
Jonan: Like, it's a huge risk. You've got a 20,000-person company, many billions of dollars a year in revenue, and you're just going to start using this thing that four people maintain out there on GitHub. That's scary to you for a lot of reasons, not because of the software quality because obviously open-source projects are fantastic but because you want someone to have your back. If I'm making that kind of shift with that many customers' livelihoods, I want to make sure that there's someone I can call at a moment's notice. So Azure adopting a lot of these open source technologies means that there's someone to make a phone call to when things go badly.
Steve: That's spot on, and I'll give you two examples. So I was working on a project in Kubernetes; specifically, AKS was involved, and we started looking at service mesh options. And then it was like, okay, what's out there for service mesh? And yeah, you have Linkerd, you have Istio. And then it was like, oh, HashiCorp has Consul. And then it came down to well, okay, we have to run our workloads in production. Who can we get support from? Who can we buy a support package from? Can we get it through Istio or LinkerD? Who can we go get support on that front? Where does the buck stop? And so it came down to like, okay, we can actually pay HashiCorp for support if support is needed. And by the way, Azure has got a managed service offering of HashiCorp Consul. Those things were factors in the decision on what service mesh to go with for an enterprise or a business that thinks about where do I go when everything else fails when my experts can't figure it out? Who do I go to next? That's one example. The other example…it actually slipped my mind, but that service mesh example is the perfect example of part of the decision-making process when companies and enterprises specifically are looking at open source solutions.
Jonan: Enterprise adoptions of open source technologies are now flying along. I think the adoption of Kubernetes is also flying. I don't recall a technology ever-growing as quickly as Kubernetes has grown. It seems to me to be almost unstoppable at this point that this is the way things are going. I know some people have opinions; one way or another, software people tend to have very strong opinions. But Kubernetes has effectively won the day, and the CNCF ecosystem has effectively won the day. So enterprises, from my perspective, are forced to adopt or die today. It's that you're not going to have a choice, really. Over the next five years, you no longer get to keep dipping your toe in cloud and be like, “I don't know. This is risky. Let's stick with these COBOL machines. Let's keep the Fortran running that we've got over here in the data center.” You got to come with us, or you're going to get left behind. Your competitors are just going to sweep you because you can't write software on a three-year cycle anymore. You can't ship every three years when your competitors are shipping three times a day.
Steve: You're spot on. And I think Kubernetes has won, and I think organizations know that they have to look at it. I think what's happening, though, and I think maybe Kelsey Hightower said this, Kubernetes at some point, is going to become a commodity. And I think a good buddy of mine, Mike Pfeiffer, has talked about this a little bit as well. Kubernetes will just become a part of the stack. The complexities of it at some point will be abstracted away from us, and I believe we're already starting to see that. We're starting to see that through things like GitOps. So GitOps is like, if I'm a developer, I can just deploy my code and work with GitHub, which Microsoft owns, which is crazy to think Microsoft owns them too. I could just commit my code, and then if you have GitOps set up and set up properly, things will just deploy in the environment where they need to go. And for those that don't know what GitOps is, it's an operating model essentially for continuous delivery to your Kubernetes environments or your cloud-native environments. So it helps you enforce a desired state and makes Git the source of truth.
So when I think about where GitOps is eventually going to go and where it's going to get to, I think of a product called Heroku, which is owned by Salesforce. If you're familiar with Heroku, working with Heroku is phenomenal. It's like, oh, I have this PHP application, or I have this Go application or this Java-based application. And I don't really care about any of the infrastructure; I just know I need to run my thing. [laughs] And literally within five minutes or less, you can have your thing running and everything you need, load balancing and everything ready to scale all of that, super easy. And I think GitOps as an industry and a community, we need to take GitOps and make working with Kubernetes for the developer the same experience that you would have going to work on Heroku. And once we get to that Nirvana, then it's like, yeah, Kubernetes is there. It’s powering a lot of things, but it's more like a commodity. You should know how to work with Kubernetes and how it works. But even to this day, do you really need to know how to deploy Kubernetes clusters from scratch on bare-metal or even VMs? No, probably not. You're probably using GKE or AWS offering or AKS. Like, let's offload that administration as much as possible to Microsoft, or Amazon, or Google.
Jonan: Yeah, and we will. And I think that that's part of why Kubernetes has been so successful is a lot of people were trying to trap us in walled gardens, and open-source naturally fights against that tendency. So Kubernetes opens this layer of interoperability, this layer of abstraction that means that if I'm running on Kubernetes and I have my Kubernetes cluster configured well, then it's honestly…I mean, it's never trivial. I won't pretend it's trivial, but someday we can see from here how it becomes trivial for me to move from AWS to Azure and over to GCP and back again.
Steve: Yes, spot on.
Jonan: And so then there's the commoditization of that cloud layer. And then Kubernetes itself will be commoditized and move beyond, and we'll have past technologies like Heroku. I spent a large part of my career working for Heroku, and my goal was to demo in 60 seconds. I could create an app with a scaffold and ship it to Heroku in 60 seconds and go to the website and show them that live. And those kinds of technologies are just going to keep stacking up so that the lower level issue, I mean, there will always be people who need to care about those things, right?
Jonan: But as far as actually delivering the software, we're freeing ourselves up to do the creative, very fulfilling part of the work, I think, and less of the fiddling with the knobs and dials that people frankly are not as interested in. So with that in mind, what do you expect is coming in the future? We have a tradition on the show of asking people to make a prediction so that we can then go back and look at our folly a year from now and see how we were right or wrong. And I think if I were to make a prediction about Kubernetes or the commoditization of cloud, that's pretty well laid out. I think most people are in agreement. But what do you think beyond that might be coming for the internet or the software industry at large over the next year or two?
Steve: So two things that I'm really excited about, and I don't talk about these things a lot, are AI and how that's going to impact the world as well as blockchain. And before everyone's like, oh, here comes their crypto talk. [laughs]
Jonan: You said the B-word. You just said it right out loud.
Steve: I'm talking like, let's think past crypto. Let's think past Bitcoin and Ethereum and all this stuff that's happening right now. There's a lot of buzz and hype, and people love it, or they hate it, whatever. There's a lot of change coming there. I don't think anyone can deny that. I'm talking about think beyond that. Think about IPFS. Think about Ethereum’s naming service being a cousin to DNS. When have we ever had anything since DNS that could potentially replace DNS?
Jonan: And DNS has needed a replacement, honestly, with the number of security vulnerabilities that were found with that.
Steve: Right. But have we had anything that really could do that? And now we have ENS with Ethereum. And I'm thinking about smart contracts that Ethereum enables and what that means for the future. And I'm thinking about NFTs and what that means. And I'm thinking past that artist that sold his digital art as an NFT for some ridiculous amount, in the millions, and all of these crazy NFT stories you hear I'm thinking past that. I'm thinking of the day when you go to sell a car, and you don't go to the DMV to transfer the title. You just do it digitally over some blockchain using the same process as an NFT. I'm thinking when you go to sell a house (and I'm selling a house pretty soon), I still have to physically go in and sign some documents. I'm thinking about the day when we no longer do that. It's sort of we do it digitally on a blockchain, so it's secure, and it's cheaper. I won't have to pay all these title fees to the title company anymore just to process my paperwork. Let's just make it happen on a blockchain and charge me like $30 and let's be done with it. I'm thinking that far in the future.
Jonan: Yeah, I just want to go and find you a key and sign it. Yeah, I think it's coming. I absolutely agree with you.
Steve: And I’m thinking about IPFS and decentralization of the internet. Just think about that for a second. I have a hard time still wrapping my head around that. What does that really mean, and what does that enable for us? How is that going to change society? Couple that with 5G, and AI, and moving faster with cloud, and things like Kubernetes and GitOps, and all this stuff. It's going to be a different world in the next ten years. I think we are at a point with blockchain like the early days of the internet.
Jonan: I think a lot of what we're seeing today in terms of resistance is perfectly valid, though. It is worth pointing out that, yes, this proof of work technology that started off in the cryptocurrency and blockchain space -- And to be clear, blockchain is a technology that extends far beyond Bitcoin and cryptocurrency. You're effectively using that to drive growth in this technology. But it's like saying PayPal is the internet.
Steve: Right. [laughs]
Jonan: PayPal is not the internet. The internet does a lot more than PayPal. And we're seeing a place now where we may be able to fix some of those technologies. It's certainly repairable. We're doing a lot of work today unnecessarily, and we can look at things like Proof of Stake as cheaper, less environmentally unfriendly ways to do this kind of blockchain technology. But no one's going to stop this train that is now a three trillion-dollar industry.
Steve: Not at all.
Jonan: We can sit here and lament the fact that it exists, or we can set about fixing it and finding ways for it to be sustainable and try and steer the technology towards powerful uses that are good for our communities and good for society in general. The future of blockchain is absolutely going to shake up the internet. We've potentially replaced DNS; we’ve replaced the file system; we’ve replaced the compute component. And while the layers of abstraction are missing, it is a very reasonable thing to expect that sometime in the next few years, we'll have a platform as a service like Heroku running on blockchain. That is a game-changer. When I can just ship my app outside of any corporation's control, really, it's going to shake up the internet in pretty big ways.
Steve: And tying that back to Microsoft, Microsoft recently announced they're killing off their blockchain service in Azure. But think before that. So they built a service called Azure Blockchain Workbench. With ABW, you could go spin up an Ethereum network in minutes, and then you could start working with creating a DApp with these examples. It would take you 15 minutes, maybe less if you knew what you were doing already. And then boom, you had a smart contract where you could process transactions, right?
Jonan: Yeah. And we're already seeing places where people run them effectively as Containers. I can ship a React app on an Ethereum blockchain. It's going to change everything. I agree with you. I hope we are able to find a way very quickly to make this technology safe and sustainable. I think a lot of the public pushback against these things is going to help, but it's not stopping.
Steve: And since we're still in the early days of it, there's an opportunity to make it run more efficiently from a power consumption perspective and make it more environment-friendly. We should be looking at that as much as we're looking at what we can do with the technologies. And I think a lot of people are…I think there's a lot of efforts out there of people trying to figure out how do I do blockchain more efficiently, use less power? It's going to change. I'm really excited about blockchain in general and what it means for the future. I'm excited more about what's coming with it in the future, more so than what's happening with it right now.
Jonan: I am too. I think there's a reasonable chance that I will not have to go down to an office and sign a paper title when I sell this home.
Jonan: And I will not send a single fax unless it's fax over blockchain. If there's not an Ethereum fax company yet, there probably will be.
Steve: Yeah. COVID-19 and what happened with the pandemic is just accelerating all of that. When that hit the U.S., and hit the rest of the world and everyone had to go work from home, enterprises, and companies, in general, were like, okay, now we need to figure out how do we make this work with people working from home? How do we digitize more of what we do and how we do things? And it's just accelerated many, many things. And I think it's accelerated some of the stuff with blockchain as well. It has made companies start to look at it like, how can we use that technology? If another pandemic comes or if we want to let everyone continue working from home, how can we use this technology, too, to maybe change the way we do business? Right now, when I go sell my house, when I go to the closing, I go on one day, and the buyer goes on another day; that way, there are less humans. We're not in the office together. So it's already happening. The next step is how do we take this and make it digital 100%? No one has to come in. [laughs] I think blockchain and smart contracts are part of what will power that.
Jonan: Yeah. If you were starting out in your career again today -- I'm sure that there are many people here listening today who aspire to be in your shoes; what advice would you give to that person? If you were to start again but in today's climate, what do you think people should be paying attention to?
Steve: I think I would do the same thing I did before and what I mean by that is go wide and then go narrow. Learn as much as you can about as much as you possibly can. Get a variety, get a good solid foundation. Once you feel like you're in a good place there, then start to specialize in an area or two. And as you specialize, never get tied to a technology to the point of if that technology dies, you feel like your career is over or you can't shift. I've had to shift what I do and what I focus on since I've specialized several times, and I'll probably do it again. More recently, I've been really focused on Kubernetes, and Containers, and DevOps, and cloud. Maybe in the future, I won't be working with that stuff as much, and I'll be more focused on blockchain; who knows? But if I have to make that shift to survive, I will do that. So as someone coming in, get as much as you can as a variety and then at some point specialize so you can get really good at what you're doing. And then don't get tied to the technology and just continue to evolve. We're in an industry where things are changing all the time. And the change is accelerating, and so you have to be ready to shift. Going back to Microsoft, if Microsoft didn't shift, if they didn't make that massive shift, and they're a big ship, if they didn't make that shift, I don't know if Microsoft -- they might be on the decline right now, to be honest.
Jonan: Oh, absolutely. I think they would be; I think it's really good advice for the next generation coming up today, the go wide and then narrow. I think it's hard to convince people early on because very often, you're coming into the industry, and you're looking at all of these technologies, and you're trying to pick the right one. The real answer is there is no right technology. You're going to choose so many things to learn and to work with over your career. Go wide, find something that interests you. That's going to drive more learning than anything else you could possibly do. You find a piece of it that just resonates with you eventually. For me, I remember starting out, and someone said, “Maybe you specialize in something like databases or cloud,” and I was thinking that those both sound terribly boring. No, anything but those things [laughter] of course, now here I am. And I work with a lot of data, and I work in the cloud and on and on. It's hard for you to predict what's coming, so just try to be prepared for the inability to prepare by going wide and learning a little bit about a lot of things and then focusing on the things that really excite you. So I want to thank you very much for coming on the show. I just want to give people an opportunity to find you on the internet. Where would people go to look you up and your work?
Steve: You can go to my blog; it’s buchatech.com. And then my Twitter is @buchatech. So, super easy.
Steve: Same thing, actually.
Jonan: Good job.
Steve: So all my stuff is just under that. So if you Google that, you'll find me.
Jonan: Excellent. Yeah, this is an important part, too. If you are listening and starting out in your career, name yourself the same thing everywhere. Choose something to the best of your ability that is globally unique and identify yourself on all of the major platforms. And as new platforms come, even if you're not going to use them, just go there and register. It's really easy when you go to find your tech friends, and they have the same handle everywhere. So that'd be great. All right, well, we are going to wrap up for today, but I thank you again very much for coming on. And I look forward to chatting some more with you a couple of years down the road when things are drastically different. It's going to be an exciting ride either way. Thank you.
Steve: Awesome. Thanks for having me.
Jonan: Thank you so much for joining us. We really appreciate it. You can find the show notes for this episode along with all of the rest of The Relicans podcasts on therelicans.com. In fact, most anything The Relicans get up to online will be on that site. We'll see you next week. Take care.