 from Austin, Texas. It's theCUBE, covering KubeCon and CloudNativeCon 2017. Brought to you by Red Hat, the Linux Foundation, and theCUBE's ecosystem partners. Okay, welcome back everyone. We are here live in Austin, Texas for KubeCon and CloudNativeCon 2017. This is theCUBE's exclusive coverage. I'm John Furrier, the co-founder of SiliconANGLE Media. It's Stu Miniman, analyst here with SiliconANGLE Media. Next guest, Justin Garrison, co-author of the CloudNative Infrastructure book. Alvin has wrap up day one with two days of live coverage. Justin, welcome to theCUBE. Thanks for having me. Stu, day one wrap up guys. What are you seeing? Justin, what's your perspective? There are a ton of announcements today. It was kind of crazy. It's amazing being part of the CNCF community and everything, and everything happens in the open, but then there's so much other stuff in the ecosystem that just happens and gets announced. I mean, real accelerated growth. If you look at KubeCon, this is the second year doing the event. Last year was kind of an inaugural event. The first year before that was just an idea, side breakout kind of thing going on, just forming and then just took off. Obviously containers with DockerCon and container ecosystem kind of floating that boat up. Kubernetes, I mean, the tracks are huge. Agenda looks like it's like a university of geeks here. I mean, this has been ramped up pretty fast. What did you take on that? I mean, I was at KubeCon last year and about 1,000 people was a great little environment and a lot of stuff that was still emerging and being discovered and now it's, everyone is in the middle of it and trying to learn as fast as they can to pick up these projects and see how to actually make this stuff production ready and how to actually use it. And not just in the environments that they used to be, they're rewriting all of their environments. It's no longer a, here's how I run my app the old way in a VM statically. It's, you know, we're beyond running containers and they realize that's not the end goal anymore. Justin, when we were talking to you before the segment here, you're like, oh, I've been working on Kubernetes for like two years more and everything and it's funny in the career. It's like, oh, well, two years in some ways, it's a long time in cloud native time. It's a long time in career wise. It's a rather short time. It gives a little bit about what you've been seeing, you know, what's interesting. You mentioned there's a whole lot of announcements. I mean, obviously new projects spinning up but new releases. So, you know, it gives a little bit of a historical view. Yeah, I mean, from last year, it was really hard to get a cluster. It was something that was really like you had to know what you were doing to make Kubernetes run and make it highly available and production ready. And now, cloud providers give you a button, you click it, how big do you want it? How much do you want to auto scale it? And it's all about the application and bringing business value to whoever's running it, to say like, my application runs here and now there's more involved with Istio and these network proxies that give you more resiliency in these cloud providers because people don't run their own infrastructure as much anymore. It's all in a cloud and they don't care about the underlying infrastructure. They care about their apps. Yeah, so you titled the book Cloud Native Infrastructure. So one of the things we're teasing out here is, you know, Kubernetes, all this cloud native stuff to make it easy to be able to do the application. But you know, infrastructure, it matters. I love Dan Cohn's line and the keynote this morning is, you know, it's exciting times for boring infrastructure. So, you know, talk about that layer, what's important about infrastructure. Bring us in a little bit, you know, why you wrote the book with Chris Nova and, you know, how that fits. Really wanted to write it to help people not make a lot of mistakes, to help them kind of level up and get a head start in building infrastructure in the cloud because it's not something that they own anymore. It's not this server that they rack and they take care of it for years. It's something that comes and goes. It's quick. You know, you have to make, you have to design for failure and resiliency and that layer of infrastructure isn't important because you don't run it anymore but it is important to build a platform on top of that that your applications are still resilient. You can't have, there's no more scheduled down times in the cloud. Like web services don't, like websites aren't gone, you know, Sundays at midnight. Yeah. It doesn't have to be pretty brave because you don't own it anymore. But if something goes wrong, you know, you're the one who drops the line. Yeah. Justin, talk about the feedback that you might have for the industry. Because Stu and I were looking at it, look at the growth. Certainly we love the excitement, but they're still running as fast as they can. They're pedaling as fast as they can. They're trying to introduce them all these services. You see some good news here, some new releases coming up, some key services we're monitoring, tracing and whatnot. But what do they need to do better? In your opinion as a practitioner, someone who's out in the trenches, what's the critical analysis? We're in a good way, constructive criticism. What needs to happen? Right. Moving forward, I mean, there was this, you know, since configuration management, everyone said infrastructure is code. And really we need to level up to be, Chris actually coined this to me in the book was infrastructure is software, where it is a piece of software that's running that you declare that has a two-way relationship. It's not a get repository that statically defines things. It's a declarative thing that mutates the infrastructure and talks back to the user so that things can auto-scale and have same defaults. And you don't have to do every last little piece of it, but the declarative nature and policy-based roles and all the infrastructure you're building and everything around your application and with your application need to be defined and controlled by software and not people anymore. And what needs to happen to make that? Obviously, SDN, Software Defined Data Center, you've seen a lot of that going on at the network level storage now. Are they there? What's the progress meter on that? How would you peg the progress of the evolution making that happen? Because that's really what people want. I mean, at the end of the day, that's what Lambda is for Amazon. That's what serverless is. That's what virtual cubelets are for Azure. Yeah, and it's funny because you can run Lambda in a very non-cloud-native way. And you can have an individual deploying code to Lambda and not checked in to get anywhere. And you can do all those things. You can go against the cloud-native model very easily. And so it's interesting just seeing that evolution as well of people actually adopting how Netflix has been doing it for a long time, where it was- We should call it software-native. Yeah, we just coined the term in the queue. I mean, but this is kind of what we're getting at. I mean, people on the buyer side are looking at it. Okay, I get it, I see the new wave coming. I want to get out there, I want to ride this. And they want to kind of vet out who's pretend or who's the player. How should businesses evaluate the pretenders and the players in your mind? A lot of it, I mean, moving to the cloud should be an easy sell. Like people building data centers anymore, you have to have immense scale to really care about those things and really get any sort of benefit. If you can beat Amazon and Google and Microsoft at the pricing game, then you're still not ahead because you still have all the people management. They have so many thousands of people that are doing this stuff that you can't keep up. And so, I mean, just adopting one of those clouds and not worrying about the vendor lock-in. Like, yes, Kubernetes brings a lot that you can move from cloud to cloud, but really it's about moving to a cloud provider that provides what your application needs and at the rate of innovation that you need. And if you can match those two things, if you can stay innovation matched with, I mean, Amazon is probably going to pull ahead of you because they're doing this as their job. But I hear you correctly what I was hearing is then, look for people that are players that are constantly introducing more innovative services. I'd be an indicator. If that's what you need, if you do not need a high rate of innovation, if you have a lot of policy or a lot of rules and regulations around your industry, then you probably will get lost in Amazon and they'll move ahead. They'll move too far for you. And so you need to find what matches for your industry and your applications. 4,000 people here, over 4,000 people here. For those that didn't come, what are they missing? What's exciting you the most? Is it the hallway track? Is it some of the special interest groups? What are you excited that you've seen so far and looking forward to seeing? Really missing, I just love the community. I mean, every talk is recorded if you really care. Like go watch them on YouTube, they're great. Like every one of these things is fantastic, but engaging with people, meeting face to face and then finding all the, because there are a lot of people that aren't online. Like I mean, on Twitter or on social networks, thousands of people here aren't. And you get to meet those people and find out what their struggles are and what they're working on and then learn from them either where they're headed or where they were before and a lot of that. There's no BS culture here. I mean, you can meet the people who write the code and they're going to give you the straight scoop or tell you they don't know it. It's real authentic. Any things that you're hearing, kind of what's the buzz, what's the pain points? You know, what you're hearing from the community so far? A lot of people still aren't in cloud. They're still doing it themselves. Standing up a cluster on Prem, you know, has struggles. They're a Kubernetes cluster that is. I mean, you can't really adopt some of these patterns until you have an API that declares all of your infrastructure. And that's still hard for people. And OpenStack was going to bring some of that stuff and sometimes it did, sometimes it didn't. But really it's about people and processes and getting those things right and being able to change the culture of your environment and free applications. That's what's important for the business and that's where you can learn from people face to face and actually talk to them and not just read a blog post about it and just hear a one-way conversation. You know, one way, tell me what you did. I need feedback and I need this feedback loop from whoever's doing it to say like, well, why did you do that? And that's really important in the community to learn. One of the proof points we're always looking for, it's one thing to see, you know, some of these cloud native companies. Like you see a Netflix, you see a Lyft, you know, these are companies where, you know, digital is their business. What about, you know, more traditional businesses? You know, are you seeing, are they able to make that change? Is it too challenging for them? You know, what are you seeing? The people in the culture is the hardest thing to change. It always is. And if they can change the people, the keynote today, they were talking about Netflix was talking about the tools influence your culture and if you can influence your culture positively and do that intentionally to actually change the people, then yeah, absolutely. They totally can pivot and make that change. But again, do they need to? I mean, if there's other government restrictions or something else that like, they could move too fast and cause other problems for their industry. What's the practitioner dream scenario right now out there that you see? Cause you made a good point. Sometimes you might want to have more services, sometimes you might want to pull back. So it kind of depends on the perspective, but generally speaking, cloud-native, Kubernetes offers an opportunity. What's the Nirvana? What's the ideal use case for practitioners these days? What's the key things that need to be rolled out or on the table that should be taken advantage of? You mean as far as technology goes? Just to soak you up, whether it's technology, whether it's mindset, culture, people, personnel, package, ops. Let's see. If we can change the people mindset of how they do things, how they deploy applications and how they manage those applications, the technology would fall into place, I believe. Because the people would drive towards this way of working and then they would build those tools just naturally. A lot of times with like Kubernetes, Google was in that mindset and so they did that. They had that culture and now they're trying to share that with everyone else and then everyone else has to learn from the tool rather than the people building that. And that's interesting. The Netflix talk on the keynote was culture and tech and I think that's a real good point because you think about it to your other point. If you've got a lot of compliance issues, you might not want to go fast. You really want to move faster. You're like a fast.com or web services company. You might want to compete on value with services. Know your culture and hire right. Know what your benefit is of your application and what environment it plays in and then you can from there figure out. And that's been the struggle for the DevOps world is they're taking a square and trying to put them around whole. Everyone want to move fast but should everyone move fast? I don't know. Depends. All right Justin, well thanks so much for coming on. Biggest surprise in this whole Kubernetes movement for you just in terms of shock factor or blew you away, made you fall out of your chair. What, share some color, personal perspective. The cult, I mean the community is just humongous now. I mean we're joining it a couple of years ago. It was pretty small and things were really difficult and now I play with clusters in Amazon and Google and Microsoft and just one click button. I play with it for a few hours and I throw it away and I got a bill for like four cents. I was like, that was amazing. Like this would, this take me so long. You know a couple of years ago and in the growth of the community around that just be able to say like this is easy now. Let's level up what we're doing and working on and figure out where the benefit is. And we were talking earlier on theCUBE and we've been saying for a couple of months this is going to bring back more time for the developers to bring craftsmanship back to the development process. Bringing an artistry and an artisan kind of real software development, not like UX stuff but like really solution driven. Focus on the business application. Where is the application and the business struggle and don't worry about the infrastructure? Justin Garrison, co-author of the Cloud Native Infrastructure book. It's on the web, check it out. Thanks for coming on theCUBE. Thanks for sharing your perspective. Day one wrap up here in Austin, Texas for KubeCon and Cloud Native Con. I'm John Furrier with Stu Miniman. Thanks for watching. See you tomorrow for day two coverage. Thanks for watching.