Techstrong Gang KubeCon Paris – March 25, 2024

Wow. Everything Kubernetes infrastructure is code open telemetry. It's been a cube con, kind of white week.

You're watching Text Junk Gang. All right. We're back here last day of, uh, Paris Cube Con.

This is a special edition text Textron Gang with a very special guest. Uh, let me introduce you to Savannah Peterson. Hey, Pam.

Nice to be here. Thanks for having me out. Yeah, Savannah.

Thank you, man. Feel Very lucky you starting the day. So, first of all, It's, it's good to have like, I'm tired of doing it with just me, Mike, and Mitchell.

So the game, it's actually the, the wrong game. Yeah. We got a real gang this time.

Um, so it's Friday in Paris. It's Friday morning, a weekend in Paris. I don't know about you, but I'm staying.

Me too. Yeah. Oh yeah.

Looking forward if I have Mine on the front end, it's great. Yeah. This is the best excuse to come to Paris ever.

There's no doubt about it. Yeah. I wish, you know, maybe they should do it here every year.

I'd be Okay. Same speaking about that. I'm hearing London in Europe next.

Yeah. I believe that is. It is.

Yeah. So London's not bad. It's not Paris, but it's not bad.

A lovely city, but it will be less full of cheese and fantastic wine. Yes. Yeah.

Yeah. Yes. Yeah.

But it is better than Detroit. Yes. Well, Chicago, Except all except music.

Nothing against my friends in Detroit. No. Yeah.

Yeah. There was some good music there. It's gonna be interesting to see Salt Lake, 'cause that Salt Lake is not, I'm curious about that One.

Generally thought of As a culture. That's gonna be an interesting Yeah. Experience.

It's been in my backyard. I'm in Denver, so Oh, nice. Yeah.

But it's a world away from Denver to Salt Lake. It's, it's, uh, anyway, there's a lot of mountains between here. Yeah.

There's some mountains between there. Exactly. So, but let, let's talk, let's talk, uh, CubeCon Kubernetes.

So the first thing I wanted to talk about is it is CubeCon is this kind of Kubernetes everywhere, right? Yeah. We're seeing Kubernetes on the edge, Kubernetes on bare metal, which has always been a thing, but now it's, it like people don't raise their eyebrow at it.

They used to like, oh, you know. But now it's, it's, it's accepted. Agree.

It's, I'm wondering what are you, what are you hearing, seeing, thinking about it? Oh, it Was interesting. We were talking from talking to Priyanka earlier this week, and, and she, like many are saying that Kubernetes is having its Linux moment.

And, and I will say, you know, I started working with Kubernetes four years ago and people were still establishing their container strategy. They didn't know what was gonna happen next. Yeah.

And now I feel like we've all made a decision, or at least a lot of the companies here have, and particularly when we're talking about the AI stack, I think there's a lot of Kubernetes action To be a lot going on there. You know, it's interesting, I was speaking to, um, Scott Johnston, CEO of Docker. Oh yeah.

I chatted with him yesterday as well. Yeah. Love Scott.

And, And we spoke about exactly that. So sort of when Docker made the realization that Kubernetes was gonna be it, and they pivoted to like a very developer led focus and, and what Docker is today, it's not the container. Right.

Although it is. But, um, I mean, it is a Linux kind of moment for It, I think. Um, what does it mean though, for people out here?

Right? It's interesting. I think to that point is it's not just end users practitioners adopting Kubernetes, all the providers, everybody's running Kubernetes underneath their A whatever.

It's great point, right? Yeah. It's part of the fabric of the infrastructure.

Mm-Hmm. So I, I hope that that knowledge and skill of operating Kubernetes and the complexity of it hopefully transforms to end users, practitioners using it to address the complexity issue. Because That's still, that's always been the challenge.

That's still an issue too. And it still is. Kubernetes is still hard as hell, man.

I'm sorry to say, but it's the truth. But we're seeing it on the edge in cars. I was gonna say, we've got a couple cars on the show floor.

Yeah. Mercedes is running Kubernetes over there. And so it may be hard, but it's, it's finding its way through That.

I mean, I need to pull over and add a few more clusters to my Kubernetes on my BMW so I can keep going Here on the auto bond. Yeah. How funny would that be?

Um, I mean, it's not far off. Yeah. You know, that's A little bit of truth.

Yeah, exactly. You know what time my car is doing, like, I get the message, we're doing an auto update. I start thinking about that.

Is it what Version of Kubernetes just came out, right? It's the containers That They're doing. Right.

What's going on with my K eights in this whip? My K couple cylinders, my K eights. Absolutely.

I'm waiting for the idio idiot. Like K eights. Come on on the deck.

Yeah. Ding ding. Um, let, I wanted to talk about bare metal though, 'cause I brought that up too.

Look, I remember I almost from the very first, uh, cube con that I attended in Seattle, There were people running Cobe on bare metal, but they were looked at as jerks, you know, quite frankly. Right. They were like, what are you doing?

Yeah, yeah. We're in the cloud, dude, sec. The second, these were people running, and I'm not gonna mention their name IBM, but you know, they were running co on bare metal in traditional data centers.

What was it? Soft layer that they had bought? Mm-Hmm.

That was the soft layer, you know, a cloud native solution, run cobe on bare metal. And, and, but there was, there was an audience for that. There were some people who really, for whatever reason, security government regulation want, did not want to be in a multi-tenant Yeah.

Type of environment. And they ranked, it was just more economical to run it on a, you know, a a bare metal individual physical server the Most out. Right?

Right. But cost is a big thing here, But now I don't know if it's cost that's driving it, or regulation compliance that's driving it. Performance.

Good Point. Mm-Hmm. But, you know, when you talk, especially, you find it, when I'm talking to people about hybrid Kubernete, hybrid cloud Kubernetes, uh, deployments, yes, you do get the multi-cloud folks, but you're getting that private public cloud kind of, you know, deployments.

And a lot of times that private cloud's bare metal. And now some of the bigger cloud guys are offering coupon bare metal as well. Mm-Hmm.

I, I also think the edge is a prime spot for it to get, become that way. Because the edge, you don't want a complex, you wanna reduce complexity. Right?

Right. Simplicity and scalability. So is that a place we're gonna see more bare metal, you know, inside the car at the edge of the network, you know, in the data center at the edge?

I think, we'll, all of that. So Makes sense. I think You're absolutely right.

Makes sense. Makes sense. It does make sense.

And people need that at speed. We need things quick. We no latency when it comes to edge.

And I think, yeah, I think you're absolutely right there. So let me be the spoil sport here. Okay.

So a little cold water. Tell Es what's, what's next, what's next? When did, has Kubernetes peaked?

And you know, the problem with peaking, especially when you get my age, you realize that right. There's one direction. It's inevitably it is inevitably a hill down.

Right? I think, I think that it's more of, uh, an intel inside moment. Right.

Okay. You know, Intel used to have their stickers on those laptops, and you knew that was your chip inside your computer. I think we're gonna get to that stage at, at a certain point, we're not gonna talk about it as much.

It's just gonna be a part of the products and solutions. But even Intel, and God bless them, we spoke with Arun and some of the Intel folks, they're big, big supporters here at Open Source. But you know, there is a MD today and God knows there's this little company, Nvidia and Arm make sure that They're supported.

You know, it's not, I mean, I, I grew up in an Intel monopoly for chips. Yeah. Right, right.

All chips were Intel, uh, or virtually, but For sure. Yeah. No pun intended.

But you know, is there, the issue is though, I don't see anything on the horizon that replaces Kubernetes No. Or Even competes with Kubernetes. I mean, there's the guys over here running Dapper.

Yeah. And you can't see it because unfortunately I can't turn our cameras this way. We All just Lean left a little bit.

You can can see it over there. You know, dapper was something that was out there before. Yeah.

Uh, but you don't really see it much in the market. Uh, the guys over at Rancher, which is Seus Mm-Hmm. They, at one time were looking to kind of do an alternate to Coop.

Well, They do their lightweight, right? Well, they do lightweight K three and stuff. No, they had a coup alternate and they abandoned that.

Mm-Hmm. Right. The original rancher os had their own orchestrator.

Mm-Hmm. Um, so you, you can't talk about having a replacement for Kube when you can't even point to one. It's mythical.

Maybe it isn't a replacement. Maybe you talked about the Linux moment. Yeah.

Maybe it is the Linux of that layer and it's just gonna be There. You, I think could be. That's actually what I think.

I don't think there, I, I think there was a little bit of competition as people were figuring out how to do containers. But now I think there's sort of, this is the car we're driving and it's it's Part of the stack. Exactly.

It's part of The solution. Interesting stuff. Yeah.

Well, alright. Cobe is far as the eye can see Co bejo. Is that what you see off the, off the deck?

You yacht? Yeah, exactly. Exactly.

Whenever I wish I was dock. Yes. Oh, UBA.

Alright. We are, we're live in Paris for Text Drunk Gang. We'll be back in a moment.

All right. Back here at Text Drunk Gang. We've got Mitch Ashley, we've got Savannah Peterson.

We're talking Cube Con wrap up. Who would, who came up with the name Open Tofu? You know, I had the exact same thought when I walked by their booth.

Like wasn't a branding kind of thing. I'm assuming No mean people who eat tofu came up with open tofu. Do you Think they're all vegetarian?

I think so. Really? I think that's what it is.

I know. Whatever theory. Yeah.

No, I I I, it's funny you brought it up. I I had the exact same thought. We didn't call it, uh, open beef steak.

So, you know, what is It? Open lobster, open Lobsters, you know, I mean, you know what? Yeah.

Yeah. So, hey, note, we should grab the people behind the Open Tofu project, like the maintainers. Maybe there's a story there.

There must be story. Maybe it's supposed to be like an a a meat alternative. Hmm.

Yeah. Okay. I, I don't know.

But meat substitute. But There's a sustainability play. Who Knows?

I mean, it's not, it's not the name I would've picked for it, but in any event, it's here and it's real. Yes. And it's kind of revitalize Infrastructure is code in infrastructure is code.

Mm-Hmm. It, I thought that had its moment in the sun Right. When we had Puppet and Chef and Ansible Terra and these kind of things and Terraform.

Exactly. But I thought those had become woodwork kind of stuff. Right?

Yeah. Like just, but no Open Tofu has, has given it a jolt of, of dare I say, excitement. Right about to you.

Dare. Yeah. I dare.

Go For it, Alan. We we validate it. Oh, dare.

Oh dare. Yeah. So, so it's dare what's gonna be interesting is at the same time this week we saw a, a bunch of reports that HashiCorp has been, well first of all, Mitchell's left as of December Mm-Hmm.

And that they've engaged a banker Yeah. To potentially sell the business. Of course, at the last, uh, cube con the big news was they had changed their licensing.

Right. Which kind of gave rise to the whole open tofu thing, right? Mm-Hmm.

And so, oh, you know, nature pours a vacuum and I think there was an opportunity there and go figure, we filled it with tofu, but, um, Of all things, Of all things. But that's not the only innovation we've seen here on infrastructure is code. I've seen one company, uh, they call it generative infrastructure, is code like ai.

Right. You know, making it easier to do. There's another company just, you know, a few doors, a few doors, a few booth down from us here.

We're all neighbors now here in Paris. We live here. This is a part of our lifestyle that's a community.

Those are just our neighbors. It's cool. Yeah.

Yeah. You know, and I'll look back and see their names, space lift there, you space lift also in infrastructures code play with a little bit of an AI twist and some other Stuff. I really found that interesting about their story is, is basically, well two things.

One is how do you scale infrastructure's code and think about Right. Software code. And how do you manage that on so many variations, many Environments point.

Yeah. Yeah. Think about the same, same for the infrastructure's code layer and the ability to have kind of one place to manage that policy security, all of that kind of thing.

Right. I thought the, That single plane of glass. Yeah, Exactly.

And the other thing, interesting thing, and I don't know if they're addressing this yet, is where does that fit into the work DevOps workflow pipeline? Right. Because That's a great question.

You know, Versions of that window application dependencies, infrastructures, code dependencies and all of that. It speaks right to the gi, GI ops, right? Mm-Hmm.

That we talked about for a long time. But that doesn't solve it all. Just putting it into the repositories.

So some interesting problems. I think it's still to solve Challenges every day. Yeah.

Well, I, I, you know, to me that's infrastructure is code at scale. At scale. Yeah.

And that's really, maybe that's the, I think that's the next moment when you're talking about that little hype blip that happened before. Yeah. I think it was, uh, uh, a marker, uh, of, you know, letting us know that this is where the market was going.

But I feel like now to your point, it's, it's vital. It's a part of the Deal. Absolutely.

I, I get it. Um, I, it'll be interesting to see how it plays out, not just from one Q content to the next, but in the market. Yeah, exactly.

We'll, we'll see how it goes. And, and again, I think if it's successful, it'll fade back into the woodwork. Exactly.

Right. Yeah. Become part of the fabric Of what we do.

Yeah. It won't be such a topic. It'll just be how things are done.

Mm-Hmm. I think we'll also have to, to see, I think this Hashi, so, so let me just say up front, right? I was, I was a big fan of Mitchell, right?

Mitchell Hashimoto. I don't know how much you know or how Well, you know, Mitchell Not, not, not as well as you do. I would suspect.

So Mitchell Hashimoto started HashiCorp, I think he was 18 years old. I mean, he was, He's impressive. He was a kid.

Yeah. And they basically built two or $3 billion product lines. Oh yeah.

Right. Uh, Terraform. B one.

Yep. Fall Mm-Hmm. Vault a whole bunch of stuff.

And there was the, the first one, not Viper No, it's the environ. I know what you're talking about. It's A, it's a security related one.

It's, um, whatever. He built at least $3 billion product lines there. It's amazing.

Fantastic. Now he's out. And, and Mitch, you and Mitchell and I have been in venture-backed startups probably as long as you've been alive.

I'm, I'm afraid to say, you know, three years that Rolled down a few hills. Yeah. That they say, you know, when the founder leaves, it's a very pro precarious time.

You know, who's going to be the, You need different talent level. It's a cultural moment. It's a product moment.

There's a lot and, and a business moment and A leader. Absolutely. Exactly.

And at the same time, it's a public entity. Yeah. Yeah.

Public company. So I, I think that's gonna be a great story to watch too, because Terraform look open. Tofus great.

They forked Terraform. Right. I'm reminded, you know, Mitchell and I were in the security game back in 2004, I think it was 2003, Ron Gula and Rena Dre.

Mm-Hmm. Reisen over at Tenable. Closed source Nessus.

7 on. Right? Mm-Hmm.

And said, Hey, it's still free, but it's no longer open source. And everyone said, oh, that's the death. Now what a stupid move by them.

They'll, you know. Yeah. It's gonna be the end of Nessus.

Well, I, you know, I'm friends with Ron. He's done pretty well for himself. He's done all For himself.

Yeah. Not too sha You know, and, and so they were wrong. Right.

It, it turned out that not being open source wasn't important. It may very well be the case with Terraform that could be that this is a That's a great point. A Fireworks moment for the tofu.

Red Bank Hats hasn't gone away and they've changed their models. Yeah, Exactly. So it, I I think that'll be interesting to see too.

Is does Terraform remain dominant or does Open Tofu continue that momentum into this whole infrastructure as code? Hopefully we can talk about it in Salt Lake. We'll see, We're gonna need something to talk about Salt Lake.

Yeah. I'm sure we'll find something because there's not gonna be nice restaurants with wine like there is here. I know.

Um, anyway, maybe we should just keep K coupon in Paris every year. I'm All for that. I a hundred percent agree with this.

It's the next three to Five. That would be seems If you're listening, if do it, we are here for it also. Is, this is the most time you've said tofu in a broadcast before.

Absolutely. Yes. We should like, well, we can't 'cause it's morning here.

But I was gonna say, we should be doing shots every time we say tofu. That's Home lit. Bars are closing at home, so we can Still do it.

Yeah. It's actually at home. You're not wrong.

You're not wrong about that. Alright. Hey, we'll be, we're, we're going to, we're gonna take our tofu.

We'll be back in a second. Our Textron gang. Hey, we're back here.

Paris. Text drunk gang talking tofu infrastructure is going Whoa. Geez.

Exactly. The food that people would talk about in Paris is tofu. You don't come to Paris for the tofu.

But, um, no, seriously. Our, our next block, our next topic for gang today is, uh, what is our next topic, gang for Gang Today it's the Open Telemetry Project. For those of you who don't know, open Telemetry is the second largest of the hundred and 80 projects.

Mm-Hmm. In CNCF. Yeah.

Wow. So, and, and in, if you counting lot of contributors. Yeah.

It's Kubernetes. Yeah. Open Telemetry.

You know who number three is? I don't. I Don't.

Argo. Oh, Argo, that Makes sense. Oh, that makes sense.

Argo, we should have known that. Coming up with a Bullet. I was holding that in my back pocket case.

You know, we Just wanted Allen to look good. You wanted it to be open Tofu I drink up. Yeah.

Next an insult. That'll Be changed by then, But No, but Open Telemetry is, is huge. It has an huge army.

Literally an army of contributors. Truly. Um, and not only that, it has that along with Prometheus, which is probably in the top five.

Yeah. I was as well have, have become the defacto foundation of an entire ecosystem. Yes.

There are very few observability vendors today who aren't on that open telemetry. Uh, very few. Yeah.

The ones that are probably stuck in monitoring Right. Or Something. Yeah.

Yeah. We still call, they don't call themselves. Yeah.

But if you think about it, that is the ultimate open source model, right. For a foundation. Totally.

We, we are, we are gonna lay the foundation, we're gonna lift all the boats, and then people compete from there. You know, we're giving you this base go forth and multiply. Right.

Go, go do your thing from there. And, and we're seeing a ton of innovation. Mm-Hmm.

But all built on that base. Well, You know, to the, to the model. And it's sort of the ideal open source project because the contributors all recognize that's not our value.

What we do on top of that is Yeah. And they kind of did that again in a smaller way. 'cause they, a bunch of different vendors are doing this profiling capability.

Right, right. That's a great point. Yeah.

How do you tune, how do you know what your performance characteristics is gonna be of your app if your stack Mm-Hmm. And they said, why are we, well, nine or 10 of us doing this, let's set that back into, into Tel and then at our analytics on top of it. So they're continuing that model, recognizing there's more opportunities to say, we're not gonna differentiate there, we're gonna differentiate what we do with that data.

Yeah. No, I think And, and it, and it's having a moment. We, we were talking a lot about it in Amsterdam a year ago.

Yes. Uh, very hot topic. But I feel to to your point, I mean, we're at, we're at an adoption peak with some of these things where it, it is becoming ubiquitous just like Kubernetes to a degree.

Well, and and I think it almost is the eng. So again, having I agree. Having done KubeCon for a long time now, almost nine years.

I bet. Um, not counting Covid, but you know, it used to be these events were very dev centric And very Kubernetes only centric. Right?

Yes. Yes. Amsterdam last year was the first time where I really looked around and said, you know what, it's about ops.

Mm-Hmm. And, and it was agreed Open telemetry and observability that I think was really kind of leading that charge. Mm-Hmm.

I think open telemetry is the engine or the locomotive pulling the train. Ooh, I like this analogy. You like that one?

I Like that. Yeah. It's a picture.

No good. You know, the locomotive pulling the train of ops and observability within this whole ecosystem. Right.

Because without it, you don't have that. You don't have that. You'd still just be doing Debts.

You have one of the leaders, Splunk, whose acquisition just got closed And finished. Right. Cisco.

Yeah. The Cisco big, Big Little already successful in observability. Now a big acquisition.

So, you know, they haven't gone wrong by going hotel. It's been a spl. Now look, there are, so Splunk is obviously a heavyweight, but Grafana or is down the block here Was gonna bring up Yeah.

Yeah. They're all, I mean, there's Some, there's some players, there Are players in there. New Relic is is using it now.

Absolutely. And they're from, you know, new Relic is from the old guard, the a perform days. Right.

The community around Open Telemetry is really impressive too. I mean, those contributors, everyone's so excited. It, it, there's gonna be maintainers there.

So I feel like there's a lot of trust in that project as well. There absolutely is. There's, um, what was the company, ServiceNow board?

Light. Light, LightStep, LightStep. I mean, they're, they're all, because they were kind of the, you know, the initiating the New writer of some, uh, distributed tracing capabilities And Yeah.

They put tracing in it. It's interesting stuff. It bodes well, I, you know, I mean, I I said in half just before about the 180, uh, projects and open in CNCF and, and they're not done.

Mm-Hmm. We interviewed Arun Gupta. Yeah.

And Arun. Yeah. Yeah.

And he's the chairman of course, of CNCF. You can catch that out. You could catch that, uh, interview.

But he said, you know, their philosophy is let as many flowers bloom, and I have runny nose here, but, um, are Your allergies respecting Those flowers? I know what Yeah. That is that what it let a million flowers.

It's great time in Paris. Yeah, exactly. So, but is it C NCF two 50?

Is it C ncf F 500? Is it the ncf FA thousand? I don't know.

Right. I've been wondering about this too. Yeah.

Is there a never, is there a limit where it's too many projects and, and so you'll have open telemetry, which is gonna suck a lot of Oxygen monolith to a degree. Yeah. Versus I don't want to say tofu again.

No. Yeah. Versus another one, you know, another project.

Right. That maybe doesn't get enough oxygen. I, you know, that's something to think about too.

Yeah. It's, we'll see, I don't know what's gonna happen. Yeah.

It's, it's a successful model. Yes, it is. Can it, Can it succeed at the next level of scale?

Kind of the same question about, and doesn't, and everything Else doesn't. Is it meant to be this size? Oh, you're good.

Yeah. Well, but, but to your point on that sna look at like CDF. Yeah.

Right. The CDF. And we do, we do a, a video series for them continuous delivery month called CD pipelines.

Nice. Fun. And, you know, we do our best to support the CDF.

Yeah. But they're having, I mean, quite frankly, they have a hard time. Yeah.

They have a hard time getting out from the shadow Yeah. Of CNCF. It, it happened even with open interest.

I believe that. Yeah. So Yeah.

We had that conversation yesterday. Yep. Yeah.

So does, does the, the, the success of CNCF itself deprive the other daughter, sister foundations of, of oxygen. Right. And, and so that's why they keep putting more projects into CNCF than these other foundations, which only kind of, you know, the rich get richer, the poor get poorer, but also puts more strain on CNCF.

I wonder if, and maybe this is happening, and I just, I'm not aware of it, but one of the things hotel also did really well was projects joining together. Right? Yeah.

Distributor tracing, et cetera. We we're different projects. And that combining things, I think is part of that strength that they've had.

Yeah. So maybe there's, maybe that's the model. It isn't just you want to become the next hotel of, uh, CNCF.

You've gotta bring people with you in other projects. And there's Combined things to combine. I think that's actually a great point.

Yeah. 'cause I mean, you're looking at people independently choosing to contribute to these projects. This isn't, open source is a pretty wild thing when you think about the community in general.

Mm-Hmm. So to your point, stronger together. As long as it's all complimentary.

Yeah. Yeah. Let's collaborate.

Yeah. I don't know. We'll see.

All right. Well, we're stronger together. At least.

Thank you all have all Absolutely. Always. Let's wrap up here.

Um, it's been all kidding aside, it has been an amazing coupon. Really has Paris has been an amazing post city. I would say.

Insane coupon. It's insane. It's been crazy.

The energy in here has been the, is back Awesome. Felt this in San Diego. Yeah.

Which was right. That's what everyone's in before. No, we're, we're back.

Yeah. We're definitely back. Paris has been an amazing host.

Thank you. Paris. Yeah.

Thank you. Doing sit. Could you imagine what it's gonna be like for the Olympics here in three months?

No, I don't want to. Yeah, the cab, I'll tell you the cat coming. Are we gonna vote to sponsor and go covered the Olympics?

I'm just gonna say were first here. You pick. You were first.

I'll wear you tshirt. Whatever. That's all I can tell you.

There're Oh, I bet. I don't want to, what's it gonna be like if, I mean, is gonna be madness. Yeah, it is gonna be.

Absolutely. It's gonna be fun. Absolute delicious Madness.

Paris in July for the Olympics. Ugh. Hashtag dreams crazy.

Those grapes are ripening on the vine. Um, anyway, hey, we hope you've enjoyed our special text Drunk Gangs while we are here. State two.

Well, by the time you're watching this, CubeCon will be over, but we've been streaming live all week as well. All of those, uh, live streamed videos are available up on Tech Drunk tv. Check 'em out.

And, uh, thanks for joining in OVO from Paris. Everybody. Savannah, thank you.

Thank you. Hey, thank you. You bow.

We're outta here. Cloud native now is the web's leading resource for the growing cloud native ecosystem. com is your destination for news, thought leadership, features and webinars on cloud native architecture, Kubernetes serverless, cloud native application development, microservices, service mesh, cloud native security, and more.

Stay on the cutting edge of modern application development at Cloud Native. Now.