Exploring Kubernetes and Cloud Innovation with AWS | KubeCon SLC 2024

This is Textron tv. Hey, everyone. We're back here live in Salt Lake City for Cube Con.

You know, it's not a lot of events where you see people in ho in, in sweaters and vests and, and coats like this. It is a little chilly out. It Is a little chilly.

Yeah, it's A little chilly. But look for this South Florida boy, I'm glad I, I don't get this that often. I get a chance to wear my sweaters.

I actually, you know, I've enjoyed it. I'll, I'm, I'll be happy to go home. You Dust them out.

You, you make sure the moss bolts Right. There's no holes in them or anything, but it's all good. I want to introduce you to my friend Nathan Tabor.

Nathan is with aws. His official title is Head of Product Kubernetes, which is a mouthful. We're gonna talk about that in a second.

But first of all, Nathan, welcome to our Thanks. Fun home here at KU at coupon. Thanks for having me.

My pleasure. I'm gonna ask about your experience here, but before we do, tell people a little bit about Nathan. Yeah.

About me. Uh, so from the cold frigid north Seattle, Washington. Okay.

Um, more wet than cold. Yeah. It's, it's not that cold.

Let's not, you Know, it's raining so much. Actually, they said it's gonna put four feet of snow down. Really?

Well, that'll be good for, That's my stack People. Yeah. It'll be very good for us.

Yeah. So, yeah. Um, I work at Amazon Web Services, been at AWS for better part of a decade at this point.

And I've been working on containers in Kubernetes the entire time. Really actually joined AWS before EKS before we launched a Kubernetes service. And have, have really watched us evolve both as a, a company and also as a community in terms of figuring out, you know, how should we be running these modern applications?

How should we be modernizing workloads? And it's really funny, Alan, you know, like 8, 9, 10 years ago, people, you know, we get on stage and say things like, you know, uh, I think Andy Jassy, right? Like, in the, in the fullness of time, everyone will move to the cloud.

And you go and you're like, everyone's already in the cloud. Well, absolutely no way. Right?

I mean, we talk to customers all the time. They're still on mainframes, they're in data centers. The amount of COBOL running on Kubernetes would astound you.

Absolutely. Absolutely. I mean, that, that, but that's, Yeah, That's one of the dirty little secrets about Kubernetes, is it's not just in the cloud.

That's right. You know, Mitchell Ashley, one of our analysts, my friend Mitchell and I have worked together 20 years. He has this saying, Kubernetes everywhere.

Yeah. It really is everywhere. It's Everywhere.

It is everywhere. And so for us, it's an on-ramp, right? Like, that's, that's why I, I was really excited about this team.

That's why I've stayed here. That's why I've been a, a part of this journey at AWS for so long, is 'cause for us, this is how customers are modernizing. So when, when someone like Andy Jassy says, in the fullness of time, everyone, you know, all these workloads will be on the cloud.

The second question is, how are they gonna get there? How is that gonna happen? And so, to me, Kubernetes is, is both the, it's a community where people can innovate freely and rapidly.

And it's also an enabler for a, a, a, a whole lot of organizations to actually realize the benefits and the optimization and the opportunities that they have, both with all the, the new stuff that's, that's being created by the folks that are here at, at, at CubeCon and also from companies like AWS who are inventing the future as far as cloud and cloud enabled services. Excellent. I love that.

Yeah. 10 years is a long time. Yeah, it is a long time.

Yeah. It goes by like that. I'm here 10 years doing this.

You're not to tell me. But, um, before we jump in, I have a bunch of questions I wanna go over, but before we get into that, I want you to talk about your title. Head of product.

Head of Product, yeah. Kubernetes. Yeah.

What exactly does that mean? Yeah, so, so what it means is that I, I lead a team of product managers and those product managers help us figure out what to, I mean, of product manager is a word that's thrown around a lot. And, and some people manage programs and processes and stuff like that.

I like to think of product managers as people that help make hard decisions. So we have a lot of really smart engineers. And if, if they were just set free you to do whatever they want, they'd make some good decisions and some bad decisions.

And so then a company sees that and they say, Hey, you have a group of people and they're making some good decisions, some bad decisions, some wins, some losses. What if we have some people that help us make really good decisions all the time? And that's pretty much what I do.

That's what we do, is we, we help make better decisions. We help build products. I spend a lot of time with customers.

I spend probably 70 to 80% of my time talking to customers, hearing what they're doing, listening to their problems. And I really like the first principles approach, right? You hear someone, they say, I'm having this problem.

I want you to solve it this way. And, and really my job is to ask why. Why is that a problem?

Why is that hard? Where are you trying to go? And, and my team, our, our entire job is to help AWS both the, the, the business side of AWS the engineering side of AWS is I'll think about how do we make really good decisions?

How do we build the right thing for our customers? How do we build things that really delight our customers? How do we build things that help at the same time bring our community of both, you know, folks in open source and also all of our partners?

How do we bring that forward and help accelerate innovation together for all the folks who are using these technologies? Excellent. So, you know, you said Andy Chassis said eventually everything kind of will, Things will move, move to the cloud.

Yeah. Yeah. And you look at AWS Kubernetes is a great example.

You guys made a bet on Kubernetes. We Did. Yeah.

Turned out to be a good bet. You know, let's be fair. But you make bets on a lot of open source projects.

We do. We do. Let's talk about how do you just, 'cause you look here, you know, people can't tell 'cause they can only see overnight when you look around, there's, as far as the eye can see, almost 11,000 people is what I heard.

Yeah. Yeah. But how many booths are there?

Yeah. There's 180 projects in CNCF. Absolutely.

How do you, where do you make, how do you pick what bets, what are the smart bets? How do you decide what, Hey, this is an open source project we're going to invest in that we're gonna bet on. Yeah, We follow.

We, uh, it's actually a really great question. And I, I got asked this question earlier this week by someone that said, how did you know to invest in Kubernetes at AWS There were a number of other competing Sure. Was container orchestration systems.

I'm, how did you know Dr. Zos? All these things?

Sure. Red. Sure.

I was like, you know what? I really don't know. That's a great question.

I I, I, I was involved, but I I, I don't have a great answer and I, I, I should have known the answer. 'cause I reached out to the person who did know the answer, edited Boas. He's a, he's a vice president now.

He's been been here even longer than me. And he said, Nathan, it was really obvious customers were telling us we had to do Kubernetes. And I said, that's a, that's a really good answer.

That's it. And, and that's, that's the truth, is that, you know, when we launched EKS, we had, um, I mean, I remember at Reinvent, there was a line around the, around the corner trying to get into the session. And, and, and the entire history has been people beating down the door.

And, and it's not because we did anything special. I mean, I think we've, we've done a, we've made a lot of very good decisions. You just read the crowd, you read the room.

Yeah. We read the room. And, and that's really why we spend most of our time talking to customers, is that people told us, we want you to help us with this.

We want to use this technology. It's hard to use. And that's really the mission, right?

So for us at AWS it's, we wanna use the technology. We, we see the benefit, but it's hard to get there. And, and that's really our job.

Like, that's my entire team's job, is how do we make it easier for people to get the benefits of Kubernetes, to get the benefits of, um, you know, all the different cloud infrastructure resources. And that's really why Kubernetes came around, right? Because you have, uh, EC2 with hundreds of different instance types.

You have AWS with all these different knobs and controls, and all of those things exist for a reason. They're very good reasons. No one at AWS does something without a very good reason.

And a, you know, usually a, a a a six to 30 page paper written behind it, uhhuh even the smallest change. So all these things exist for a reason, but when you put it all together, it can be hard to use. And so systems like Kubernetes help abstract that just a little bit.

They let customers access the controls, but also automate around it. And our job is to make that really, really easy. How do you access that automation?

How do you go from, you know, tool to value quickly? And that's really what we're all about. And, and Open Source is particularly good at that.

We see so much innovation coming outta open source communities. And so for us, open Source is a place where we're gonna continue to invest. And we've seen a lot of success helping customers with open source, because we think that there's this flywheel that drives innovation outside of Amazon.

Inside of Amazon. We have a lot of teams doing a lot of, of so much different work in open source. I, I tried to memorize all the different places where we invest in open source.

And I, I, I can give you a few, but I can't remember 'em all at this point. But, um, but yeah, it's, it's, it's been this amazing part of our flywheel that drives customer success and, and acceleration of the ability for customers to innovate. Tell you a funny story.

Go to, I went to DockerCon in Austin in about 20 16, 20 15. Yeah, absolutely. We had it launched, which is now cloud native now.

It used to be called Container Journal. At the time I was thinking of naming a Docker something, but Docker, you know, I would've got sued. And I came back from that DockerCon and I told my team, I said, you know, I'm glad we're not going with Docker.

I think container's the right way to go. Ah, yeah. I said, but there was this other thing there, this thing called Kubernetes.

Everyone was talking about it. It had all the crowds. It'll never catch on.

It's too hard. It's really hard. Yeah.

And that's why, you know, that's AWS that's aws swar the crowd and said, I'll make it easier. Let's Make it easy. That's what we do.

That's what we do. That is exactly it. Exactly.

It, it's an excellent point. I got some other stuff here, Nathan, I want to hit on. Um, so let's fast forward to today now, right?

Kubernetes is a, Kubernetes is the, the engine that pulls this whole train. Yeah, yeah, yeah. In many ways.

Yeah. How are you guys involved today in continuing innovation and continuing in leadership? I mean, AWS is a leader.

Yeah. So in the, in the cloud native, you know, we're, so we're involved in a number of open source foundations in general, right? So we have both project contributions and sponsorship at, you know, the Apache Foundation, at the Eclipse Foundation, Linux Foundation, cloud Native Computing Foundation, which we are, we are, or we are right now.

Yep. So, but in particular in the Cloud Native Computing Foundation and with Kubernetes, we, we participate in several ways. So at the, at the high level, strategically, we, we participate in the organization.

So we sponsor, right? We are, we're here, we, we, we, we give money to sponsor, but we go beyond just like, Hey, we're gonna buy a booth. We actually have given $3 million in crowd cloud credits and engineering resources to the C NCF F so that they can build automation that helps all the projects that are part of CNCF do their testing, do their validation, do their hosting.

So that's really important. Us that's fantastic us to actually sponsor the project. Um, Kubernetes, obviously, we, we have a number of folks who are contributing, uh, to the upstream.

We have a number of maintainers on key projects on AWS. We have on, on, on our team, actually in the Kubernetes team, we have people who are on the TOC for Kubernetes. We have a person who's on the security response committee.

So when security incidents come through and there needs to make a decision about how do we handle different critical vulnerabilities and exposures, CVEs, we actually have representation and we have people who attend those meetings, who review project submissions and do all that kind of stuff, right? So we're, we're, we're participating financially, we're participating with our time, and then we're also contributing products. We're actually, we're, we're, we're we're doing upstream changes.

I've actually contributed to Kubernetes, which is amazing. Not a lot. Yeah.

Oh, you've contributed code? I have, I have a few, uh, documentation. Okay.

And it's still 98% of the rest of the people out there. So we, we, you know, we have folks who, who, who mer make, you know, they write caps, they do upstream changes. We've also, I've, I've been really, um, probably my, my proudest moments as a, as a part of this team is where we've, uh, given projects to the CNCF.

So I have two projects that I've been a part of giving to the CNCF. The first was the CDK for Kubernetes, which a sandbox project. Sure.

Really excited about that. And then the other one has been carpenter. And Carpenter is, we Were just talking about carpenter with op, I don't know if they're a finops.

Yeah, yeah. We Love Carpenter. Yeah.

So I, Big stuff. So my team, we help build Carpenter and for us, something like that, being able to give it to the CNCF, put it into SIG auto scaling, make it part of Kubernetes Dream come true, because we were able to develop this new innovation at AWS and then make it part of the project. So, you know, actually building this innovation, externalizing our innovation, giving it to the community, making everything work a little bit better.

That's, that's really the, uh, that's the other way that we contribute. And, and, you know, I, I think of all the ways we contribute, that's the way that I'm most proud about, about the impact that we've Had. Yeah.

I think you're a hundred percent right. I mean, money's, as my father-in-law, rest, his soul used to say, rich for, it's nice to have money, and money's always nice. Absolutely.

Absolutely. But contributing projects. Yeah.

Which it's kind of sweat, right? Yeah. Yeah.

That's And ip, Right. And ip, exactly. That's the value.

It's, it's, it's invaluable actually, right? Yeah. And it's been important for us.

I absolutely, I don't think we would ever have done Carpenter a different way. Yep. So I just wanna make sure we hit all our little things here.

We about some open source innovations there and why it's important. Um, let, let's go with this Now. Let's, you know, we'll put on our glasses.

First of all, I can't believe we didn't mention ai. Yeah. So I'm gonna tie AI into this question.

We talk about ai. Absolutely. You know, when we look forward at the future of EKS, I'm sure AI's gotta play in there somewhere.

Yeah. Yeah. But AI and everything else, what can you, without, let's not get in trouble, but what could you, and also we'll, so save two weeks as reinvent.

We'll have one there. Yeah. Well, there'll be a lot of AI at Reinvent.

I can tell you this. AI for the EKS semen for Kubernetes is, is not new. And that's the coolest thing, is that for us, we're, I'm, I'm, I'm looking at my second wave of AI on EKS.

So we, we've been doing AI and, uh, deep learning and ML model training on EKS and Kubernetes from the very moment that we launched the service. So our first wave was actually a lot of autonomous vehicle training. Um, and I really, you know, again, I can't name names because I don't wanna get in trouble.

No, No. We know where the bodies are buried. You can, you can look at a lot of the, uh, large autonomous vehicle training efforts that were going on in, in 20 18, 20 19, 20 20.

Uh, you know, a, a fair number of those were actually doing their training on EKS. And then you fast forward and you see what happened with OpenAI and Chat GPT. And they built a, a very beautiful stack that was based on Kubernetes and Ray Sure.

And a few other components. And, and so that's really been the second wave for us. And we've seen a number of foundational models for LLMs that have been trained on, on EKS.

Um, and I think we, you know, we have a good partner in Anthropic at AWS, which I spoke with, uh, with, uh, anthropic at Reinvent last year about some of the work they've been doing on EKS, that's on YouTube. You can go look it up. Um, and we have a number of others as well, that, that have actually come to EKS.

They've been scaling out and training foundational models for ai. Inference is now the big thing. There's a lot of scaling out on inference for these models.

And so for us, AI is, is, is a bread and butter part of what we do for customers, and it's been part of our, our story for the last seven years. I love it. Yeah.

We've, we've put a lot in here, but what's, okay. Um, I want talk just briefly, briefly. Reinvent is in, uh, two weeks.

Yeah. It's coming up two and a half weeks, something like that. I'm the only person allowed to be here on my team, because Everyone else, everyone studying for reinvent.

Well, we'll be out there. We'll be doing video up. You know, we have a suite at the win and everything.

But again, without saying anything you're not supposed to. What can people look forward to at Reinvent? Yeah.

Where are we going? What's the direction of EKS? Right.

I mean, I, you know, directionally, it's, it's the same story. We're not changing our strategy. Just what I told you a minute ago.

You know, our job at AWS is to make this technology accessible to customers, to help them accelerate innovation. I, I don't think our customers don't come and they say, Hey, we, we want to configure a hundred knobs and dials and hire a team of experts in order to make our codeship faster and build these new products. No way.

They come to us and they say, how can we turn CapEx into opex? Sure. How can we get to value faster?

How can I give my teams a platform to innovate on and to ship their code into production in days instead of weeks or months? Right. And so on the EKS team, that's our same story.

So what you're gonna see at Reinvent is a whole bunch of ways. And, and we have some really cool launches headed up. I Okay.

Oh man. I'm so excited. Don't, Don't, don't get us in trouble.

I won't Get us in trouble. But I'm really excited about what we're shipping, uh, this year, both at, at a, on the EKS team and across AWS There's some really cool stuff in the pipeline for customers. And, and really that's the mission, is how do we help you get to value faster?

How do we make this easier to use? How do we make it easier to scale? There's a lot of really great stuff that we're working on.

And, and, and we're not done, by the way. I mean, so to me, I think, um, you know, another, a Amazonian, you know, famous quote is, um, some analyst at some point asked Jeff Bezos, this is back in like 1998. You know what, that was a great quarter.

Are you thinking about your quarter? How are you closing the quarter? He says, I'm thinking about what my quarter that's gonna close two years from now.

Right? And so for me, we have some great stuff on the cart that's gonna come out at Reinvent, but our team is already on to thinking about, Hey, what are we gonna do in 2025? What are we gonna do in 2026?

What are the trends that we're seeing with customers around ai, around augmenting workflows, around moving faster, delivering projects anywhere they need to deliver, um, and really investing to make that customer experience better and help customers get to value both on and off the cloud a lot faster. I love it. They're closing the expo, I think is what they're saying.

So that means we've gotta close. But, um, are you presenting at Reinvent? I Am.

Yeah. You can come and Find me. Give us your plug, let here.

KKV. Yeah, so we have a whole bunch of sessions. Actually.

Really cool thing this year is the first year at Reinvent that we actually have a dedicated Kubernetes track. So there'll be, I, I can't even, it's at least 30 plus sessions that are totally dedicated to Kubernetes. We have, uh, breakout sessions, we have workshops, we have chalk talks, we have builder sessions.

I'm presenting KUB 2 0 1 Cube 2 0 1. I'm gonna be presenting with a special guest. Um, we have a, a guest from Snowflake.

Oh, Very cool. The The scale data company. Sure.

Uh, they're gonna come in and talk with me. We're gonna talk about ai, we're gonna talk about data, how to run Kubernetes at scale. It's gonna be an absolutely fantastic session.

So if you're at reinvent, uh, if you see me, stop me. I'm excited to meet you. And, and it's gonna be a great, uh, just a great time.

So it Always is. Yeah, it always is. Cool.

All right. Hey, we gotta wrap up. They closed the expo floor here on us.

Nathan, it's a pleasure to have. Thank you so much. This is the first time I've met Nathan in person.

We did virtual events together, so You did? Yeah. It's good to see You in the sweater.

It good to be in the flesh. Sweater In the sweater. You think?

It's not often. I wear sweater. Yeah.

I, I tell you, I think rare day, maybe the, You can ditch it. Well, Not you think? No, it gets cold.

When I was younger, I remember Vegas always being warm. But that's 'cause I come from the security world and the security events in Vegas are always audience. I was gonna say climate change, right?

Yeah. No. Well, there is no climate.

Don't, don't get me started. We're getting too deep. Yeah.

Don't go there. But that being said, it gets cold there this time of year. I'll probably bring a sweater with me.

You, if you, if you wanna warm up, you need to go back to, uh, back to the security world. Well, I live in South Florida. I just go home.

But I, I am looking forward to seeing the Eagles at the sphere. Yes. At the end of, uh, reinvent.

That'll, what a great way to end reinvent for me. Yeah. Excellent man.

Excellent. Thanks Alan again. It's been a pleasure.

Nathan Tabor. Thank you everyone. Head of product Kubernetes at AWS that wraps up day two of of our Kube coverage.

We're gonna be here tomorrow and it's gonna be a great day. We'll see you then. Until then, this Alan Shimel for Text Drunk tv.

We're out.