KubeCon + CloudNativeCon NA 2024 | Bobby Allen & Gari Singh, Google

Clips
News
More from KubeCon + CloudNativeCon NA 2024

Bobby Allen

Cloud Therapist

Google Cloud

Gari Singh

Product Manager, Google Cloud

Google

Three insights you might have missed from KubeCon + CloudNativeCon NA

Artificial intelligence is driving innovation in cloud-native technologies, streamlining processes and enhancing decision-making across industries. Kubernetes AI continues to reshape enterprise operations, tackling challenges from scaling artificial intelligence workloads to addressing security, compliance and modernization needs.As organizations navigate these complexities, automation and innovation remain central to unlocking Kubernetes’ full potential, according to Mike Barrett (pictured, left), vice president and general manager of hybrid cloud platforms at Red Hat Inc

Breaking barriers: Google’s GKE upgrades redefine Kubernetes scalability for AI workloads

Earlier this week, Google Cloud announced it has upgraded its Google Kubernetes Engine capacity, with support for 65,000-node clusters. That’s up from its current support for 15,000-node clusters.The jump is possible due to Spanner and other technologies, according to Gari Singh (pictured, left), product manager, Google Cloud, at Google LLC. Still, lots of the work was backported into open source.Gari Singh and Bobby Allen of Google talk with theCUBE about the Google Kubernetes Engine.“I mean, think about it. We were at 1,000, then 5,000 was big, 15,000 was massive. Now, from 15,000 to 65,000,” Singh said

play_circle_outline GKE's evolution and recent announcements, like 65,000 node support

play_circle_outline Advancements in Kubernetes simplifying networking and addressing enterprise challenges

play_circle_outline Kubernetes as a versatile platform for GenAI workloads and modernizing applications

play_circle_outline Range of customer requirements, from high-tech to simplicity

Info
Transcript

Bobby Allen & Gari Singh, Google

Bobby Allen

Cloud Therapist Google Cloud

Gari Singh

Product Manager, Google Cloud Google

In the midst of covering Day 2 of the event, Savannah Peterson and Rob Strechay are joined by Bobby and Gary to discuss their backgrounds. Gary shares his journey from working with mainframes to his current focus on Kubernetes at Google. They highlight the recent announcement of support for 65,000 nodes on GKE, as well as other updates and features that aim to simplify operations for users.

The conversation shifts to the intersection of Kubernetes and AI, emphasizing how the platform can accommodate a variety of workloads and technologies. Customers ... Read more

explore Keep Exploring

What is the significance of Gary in the Google Cloud product management team? add

What is the current state of flexibility and advancement in Kubernetes technology, and why is now a good time to start using it? add

What is the potential benefit of using Kubernetes for scaling out Gen. AI workloads and modernizing apps simultaneously? add

What are some examples of the diverse and sometimes conflicting requirements that customers have in AI conversations? add

bolt Powered by CUBE AI

Bobby Allen & Gari Singh, Google

search

Savannah Peterson

>> Good afternoon cloud community and welcome back to wonderful Salt Lake City, Utah. We are here midway through day two of three days of coverage here on theCUBE. My name's Savannah Peterson, joined by my co-conspirator co-pilot and co-host, Rob Strechay.

>> Wow, I got all three this time.

Savannah Peterson

>> I'm trying to keep you on your toes.

>> You are keeping me on my toes. I love it. I love it.

Savannah Peterson

>> In these super comfy chairs. What else do we have?

>> Yeah, super comfy.

Savannah Peterson

>> You and I mentioned that it's friends and family day.

>> It is.

Savannah Peterson

>> And I am really excited to introduce a family member and a new friend, Bobby and Gary. Thank you so much for being here today.

>> Thank you for having us.

Savannah Peterson

>> Thank you.

Savannah Peterson

>> It's always a treat. And Bobby and I have now done this a few times. And when he said he was bringing you on, Gary, the hype was real.

>> Nice.

Savannah Peterson

>> So I hope you're ready to bring it. Since the audience hasn't had a chance to meet you before-

>> Sure....

Savannah Peterson

>> through us, give us a little bit of your background. You're a total technology baller. Break it down.

>> Yeah. I mean, I've been around for a long time, so I guess technology-wise worked with a lot of companies. But I've done everything from the mainframes to desktop computing to the first range of distributed computing. As I like to say, I used Java when it was called Java and J2EE. It had two E's back then.

Savannah Peterson

>> Oh, you showing your stripes there?

>> Yeah, I'm telling you. I've been through all those. Sun was a company -

Savannah Peterson

>> Wow....

>> work with servers, so I run the gamut all the way up to everything, trading, everything. Then built a bunch of stuff on ... Got hooked on containers in 2014, I think. I was at IBM at the time.

Savannah Peterson

>> I was going to say, that's early days. That's Kubernetes inception days.

Savannah Peterson

>> Yeah. We were trying to build stuff and we had all these things to package, Java, VMs, different versions where VMs were kind of heavy. Docker looked cool.

Savannah Peterson

>> Oh, we all remember those days.

Savannah Peterson

>> Yeah, yeah, yeah. And we were building our own thing and then Kubernetes came out and I was like, "Oh, Kubernetes sounds good." Started working on that for a bunch of years and then got the opportunity ... Google called and said, "Hey, you want to work here?"

Savannah Peterson

>> I was just going to ask how you ended up on the GKE team.

Savannah Peterson

>> Google called and they said, "Do you want to come?" I said, "Sure," but I only wanted to work on Kubernetes.

Savannah Peterson

>> Oo, I like that.

Savannah Peterson

>> And they happened to have some Kubernetes jobs and for good or bad, can't determine yet which, they hired me and I'm here.

Savannah Peterson

>> Well, we're glad that you are. I'm really curious because that's a specific call-out in general. You only want to work on Kubernetes projects. Tell me why.

Savannah Peterson

>> I guess, I'm just a huge fan of ... I guess, I'm an infrastructure guy, middleware guy. So I spent, as I said on the Java side. But I'm a big fan of distributed computing. It brings everything together, network, distributed computing, running apps and things like that. Kubernetes seemed to be the base thing that was going on. I'd spent five years managing tons of clusters. And I mean, at the end, I guess I'll show ... Google's Google. It was a place that Kubernetes was born and I wanted to come see what it was like there. I wanted to try Borg, the thing that it came from. And it lived ... I mean ... They'll fire me anyway. It lived up to the hype. It was just seeing what everything was like, seeing how it works. But I really do think that Kubernetes was a thing that was going to persist. It seemed like it was at the spot where Java was at one point where everybody decided they needed Java and an app server. I'm like, even 10 years in it was like, it doesn't seem like saying that I'm going to use Kubernetes is not the wrong decision. And I was like, so it made sense to stay on that and of course I came to the place that it was born because that would be the most fun.

Savannah Peterson

>> You came to his mothership.

Savannah Peterson

>> Yeah.

>> Can I jump in, Savannah? He's being modest.

Savannah Peterson

>> Of course you can, Bob.

Savannah Peterson

>> So sometimes in hip-hop they'll say your favorite rapper's favorite rapper. He's your favorite tech guy's favorite tech guy. So he's the guy, honestly. He's being super modest, but when the GKE PMs have a question, they usually go to Gary. So he's that guy. So he's honestly, in my opinion, Gary's probably one of the widest and deepest product managers in all of Google Cloud and he's super modest about that. But if you want to have a question about Kubernetes or Cloud Run, this is the guy to talk to about that.

Savannah Peterson

>> Well, I'm glad you brought him to us today, Bobby. Speaking of the Google Kubernetes engine, big evolution here. We're at the 10 year anniversary. Where are y'all at these days? Lots of exciting things going on.

Savannah Peterson

>> Lots of big things. So if I was to give where we are a title, I would say the GK is growing up and glowing up.

Savannah Peterson

>> Love me a glow up.

>> So Gary's going to talk about-

>> Glow up. I didn't know what-...

>> the growing up-...

>> a glow up was....

>> and I'm going to talk about the glow up thing. So what are some of the things that we released recently that we might want to tell the audience about?

>> I guess the big one last night, and we have good stickers and shirts. Come on by.

Savannah Peterson

>> I saw a six, a five and on that.

>> A six, a five. Yeah. So we just announced 65,000 nodes support on GKE for super huge training workloads. We had-

Savannah Peterson

>> Super huge is an understatement, mate. That is epic.

>> -

>> We had 15K before. And then, now we're up to 65K. And as a call-out we did in open source ... So on our side, we're able to do that because of things like Spanner and other technologies. We can scale to massive clusters. But we did backport a lot of work into open source. So theoretically, somebody can run 30,000 on open source now too, which is like-

Savannah Peterson

>> Wow.

>> I'm not sure they could. Technically, there's no holes to it. So that's just ... I mean, think about it. We were at 1,000, then 5,000 was big. 15,000 was massive. And now from 15,000 to 65,000.

Savannah Peterson

>> And now you're 15,000 more than that. I was just going to say, that's a huge-

>> I think we're going ... But that's where ... I mean, we'll try not to probably talk about too much GenAI because it's everywhere, but that's where it's going. People need these massive clusters. Our internal stuff needs it and our top customers need it. So yeah, the team did a lot of work. They got T-shirts right out of it. Big announcements. But I think that's been years of work to go into that, from everything from the back end and making the scale and it was a lot of things that kind of came together to be able to make that happen. So yeah, 65K is great.

Savannah Peterson

>> Congratulations. I mean, a huge milestone and the scale and scope of that-

>> Is huge....

Savannah Peterson

>> is no joke. When Bobby was telling me, and I saw the , I was like, wait a minute. Last time we talked this was 50,000 less. This is-

Savannah Peterson

>> I think the principle though, Savannah, is ... And I want to make sure we come get the people that are maybe not doing the mack daddy training jobs just yet.

Savannah Peterson

>> You don't have to be doing 65K.

>> Yeah, yeah. -

>> You might not be doing -

Savannah Peterson

>> Just to be clear, you are-

>> It's okay....

Savannah Peterson

>> a useful ally and -

Savannah Peterson

>> You're not on the struggle bus if you're not doing 50,000.

Savannah Peterson

>> Everyone at home right now feeling a little small.

>> if you don't get 40,000 . But the principle is, we're solving the problems before you hit the barrier. So I call it headroom for innovation. So before you ... If you had an environment where we took the barriers off, we took the limits off, what could you do? That's what we're trying to do. So before you get there, we want to go ahead and break through that barrier so you can just be unleashed.

>> Yeah, I think that, to me, definitely-

Savannah Peterson

>> Definitely want to be unleashed. Love that.

>> Yeah, I mean it's huge, because especially ... I know GenAI is everywhere and all of this, but the companies are really trying to go down that path and build it into their other applications as well. AI means so many different things to so many different people.

>> It does, yeah.

>> To your point, some people are like, "Hey, I'm happy using it via API. Maybe I'm using one of yours via API." Or they're looking to go and do an SLM, a small language model or specific language model and build up taking their corporate data that they hold dear to themselves and fine-tuning and training and putting guardrails on. How do you see what AI means to so many different people and how that's really going?

>> So I'm going to answer your question in a little bit different way. So we talked a little bit about GKE growing up, now about the glowing up. Okay? So I had a really weird situation in high school. I had almost like an ugly duckling experience.

Savannah Peterson

>> Didn't we all?

>> But mine was really, really extreme. So first three years of high school I was the kid with the big glasses, Latin club. I didn't have a whole lot of swagger. Okay? And then I got contacts and car keys before my senior year that summer. Homecoming king, best dressed, most likely to succeed all of my senior year. I came back to school and folks were like, "Have you seen the new kid?" They were like, "Bobby? The kid who's already been here." So here's my analogy.

Savannah Peterson

>> Hey-yo.

Savannah Peterson

>> GenAI is the contacts and car keys for Kubernetes. So it's kind of putting a different spin, so people that are like, "Is this the new kid?" No, no, this is the kid that's been here for 10 years. And so, the glow up was a different cosmetic packaging, but there was a lot of substance there. I think that's what we're seeing.

Savannah Peterson

>> I think it's a great point and I love that analogy when we brought it up on Bobby D's podcast the other day. I think there's a lot there though. Kubernetes adoption, as you and I have personally talked about, is not quite at the level of saturation that I think sometimes people perceive it to be. We're still only seeing a 20%-ish, maybe 25% if we're being generous at enterprise, at scale. People obviously playing with that in a sandbox or new deployments, but truly at scale it's still taking time. I really do feel like GenAI and AI has put the jet pack behind that horse or that cart to really launch Kubernetes as the platform for anything you're building new, which at this point and for a lot of people's cases is AI at least minded. Not that everything has to be AI, but building those systems. It's not just that though that y'all announced today. It's not just the 65,000 nodes. Bobby, I know you got some highlights for me. You've got some self-service, you got some ArgoCD.

>> We do, we do. So one of the things, again, AI is one of the ... Kind of the shiny object that people look at, but we're also trying to get people tools to operate better. So self-service is a big thing. So we've produced some really good assets this week that sometimes get drowned out by the big announcements like the 65,000 nodes. But if you have multi-clusters and want to have self-service, we reduce some really cool stuff ... Or introduced some really cool things. Fleets teams, ArgoCDs. So, how do you allow your developers to have more self-service and capabilities based on top of Kubernetes? We produce some really cool stuff around security. DNS access to the control plane. We've got a great customer story around a customer who actually experienced the trifecta; better performance, lower latency and lower cost because of some things we rolled out. C4 VMs, custom compute classes, kind of our way to ... For lack of a better term, I guess an alternative to Carpenter, with some other stuff thrown in there.

>> I'd say, similar capabilities in a different packaging.

Savannah Peterson

>> Our flavor on it.

>> -

Savannah Peterson

>> I'm missing something. What else have we done this week that I'm missing?

>> Those are the big ones. Yeah, I mean, there was a few internal things. I mean, it's funny. There's always the big announcements that are in there and then ... But to your point, I think on where we are in Kubernetes, sometimes there's just simple things that you have to do as sort of housekeeping and whatever. I mean, as a side thing, just to call out, it was great to see the awards this morning for the folks who did a lot of work on cleaning up Kubernetes stuff and whatever.

>> Yeah, absolutely.

>> It wasn't the new features and whatever. So it's great to see them. And it's great to see us do that. We did our own little container event the other day and I just ad-libbed it at the end. I was like, "Hey, here's five cool things that you should know about. And I think it wouldn't necessarily be the coolest things to the audience, but they're simple things. How do you simplify networking? Everybody hates networking. And Kubernetes is unforgiving if you screw up networking,

Savannah Peterson

>> That is a good call-out.

Savannah Peterson

>> If you give it too small of a subnet and you're like-

Savannah Peterson

>> That web gets too crunchy and you can't-

>> You can't change stuff. But now we're making all that stuff flexible. Part of it in open source, part of it in GKE. And you're like, you kind of ... I was talking to our director of engineer and he's like, his favorite part was like ... I'm like, "What, you didn't like my great thing on our new auto-scaling and everything?" He's like, "Well, that's good, Gary," but he's like, "I liked how people smiled when we said, 'You remember how you used to have to do X, and now that's no longer a problem and the friction's gone?'" And I think that's the great thing about where we are with sort of Kubernetes, advancing and also mapping to some of these core enterprise problems. And as I tell people sometimes, I'm like, "Hey, now's a great time to pick Kubernetes." We spent the last 10 years fixing all the problems that you might've had. So we have all the answers now. So go ahead and get going. But I think the other thing with the AI, just as a ... I don't know, it's non-sequitur. But-

Savannah Peterson

>> We're here for all of it, Gary.

>> Normally how a new thing comes out and there'll be a new platform ... And I think the amazing thing about Kubernetes is that people could say, "Well, maybe it wasn't specific enough. You can run anything on it." Well, there's been a few things that tried to say maybe we should do Gen.AI and AI this way. And then, in the end, even all the frameworks that tried to build their own run times are all ... Now we've got Ray on Kubernetes, KubeRay and all this. You've got things like DRA coming, dynamic resource allocation for GPUs and you're like, wait a second. So Kubernetes, the same thing that I haven't gotten to yet maybe for my workloads but want to get to is also the same platform that I can use to scale out my Gen.AI workloads. So this is great. Now it's not this thing, like a miss out or pass out on technology. So to me, I give all credit to the folks who designed Kubernetes as being highly pluggable and flexible and I think they really proved it out. But it's like ... So like you said, I want to do Gen.AI, but now it's still the same platform that I can modernize my apps too, which is you don't have to skip a generation or technology. No confusion. Pick Kubernetes and you're okay.

Savannah Peterson

>> And it can grow and adapt with you, whatever you're deploying, whatever you're building.

>> Exactly.

Savannah Peterson

>> I do think it's really been this kind of match made in tech heaven vibe between Kubernetes and everything that's going on with AI. And it's also ... And I'm curious, confirm or deny this. This is my hunch as an ... Well, not an outsider, but as a journalist on this side, is I really do feel like it has accelerated the conversation from a strategy perspective with companies across verticals. Not just in our world or in specific little niches who happen to have been on the Kubernetes bandwagon. So you're seeing that as well? Because they're all coming to you, I would imagine.

>> They are coming to us. Yes.

Savannah Peterson

>> Has any of those conversations been surprising or fascinating from a customer perspective that you can share, obviously?

Savannah Peterson

>> I mean, I think part of what's surprising is I think people are willing to be a little bit more vulnerable now, that I really don't know what we want. And I think that sometimes-

Savannah Peterson

>> Interesting.

>> Again, Gary and I's team is really big on customer empathy and trying to meet people where they are. And so, I think we're finding that sometimes folks are still forming the question, so we can't just rush to the answer because we have to help them figure out, we can cook something, but what do you want for dinner? We might have to give you some appetizers so you can taste a few things. But I think we're also seeing that you can't revolutionize everything that you've got. So I got to bring a little Greek in here. Okay? English is super word poor and we use adjectives a lot. Like, what kind of table? What kind of chair? They're two different kinds of new and Greek, and I'm going somewhere with this. So there's Neos and there's kainos. One is new, like a new pair of sneakers, and one is new like a jet pack. And I think where a lot of customers are struggling is, do I need sneakers or do I need jet packs? You probably need both.

Savannah Peterson

>> I was going to say, I want both.

Savannah Peterson

>> You need both.

Savannah Peterson

>> You know I love sneakers.

Savannah Peterson

>> But I think sometimes, Savannah, sometimes we're mixing the two up. So if you're going to Thanksgiving dinner later this month, you don't want grandma to totally redo that macaroni and cheese. You want her to do it again like a new pair of sneakers. You don't want her to reimagine macaroni and cheese. Now maybe do skim milk instead of whole milk.

Savannah Peterson

>> Maybe a little truffle cheese on top or something to give it a little extra.

>> Cut the fat a little bit so we don't have to spend as much time on the treadmill. But the reality is, you want her to do it again, not do something brand new. But there are other things you do want to do brand new. And so, what kind of new are we talking about, is a conversation we do need to have. Like, this stuff needs to run better. Gary talked about optimization, performance, latency, but then we want to do something amazing over here. How do we balance those two things?

Savannah Peterson

>> Well, and that's where you come in to help them navigate that.

>> Hopefully. We don't have all the ... Well, he's got all the answers. I just work with smart people like him.

>> -

Savannah Peterson

>> Well, we're helping people learn how to think though and ask the right questions, because we are ... I mean, it sounds silly. Well, you're ex-IBM. The think differently side of where ... I didn't mean to throw that pun in there, that just came out organically. But I do think we're really at ... Sometimes I surprise even myself. But I do think we're at an intersection here where to your exact point, that vulnerability and that curiosity is, we're not in this kind of linear growth and scale. We're at this wild point where it's like, whoa, I want to make sure my sneakers are on too when I put the jet pack on. I don't want to be barefoot in outer space.

>> Exactly.

>> But I also think you hit on a really good thing is that people want simplicity. And I know you have a run that is the more simple version of Kubernetes, but it's built on all the technology that's under the hood. And that has to be, like you were saying, the networking thing, having been at a separate hyperscaler, I can tell you networking was the bane of my existence with anything I did at that hyperscaler. And so, that has to be a lot of what people come to you and say, "Hey, we can give you Run. This is where you can start to cut your teeth, start to do things with some of the other Linux foundation stuff like PyTorch and other things and bring in that AI there and get used to that. And then when you're really looking to scale, you can go to GKE and really get that going." Is that the kind of patterns you're starting to see?

>> I think we definitely see that. We even see people going sort of both ways. I mean, on the pure container side ... So of course there's going to be people who are ... I mean, I think the magic of Kubernetes that it did all the other things we said that it lived up to is portability. I mean, whether or not you really are going to go to multiple clouds or not or on-prem, it's proven that yes, you have to move the workloads or whatever. But from an API perspective, your stuff is still compatible. Sure, maybe storage and move stuff around.

>> There's some a little bit . Yeah.

>> But people feel like that's the sort of common denominator. So there's always going to be some who are just going to be like, "I'm not going to move away from that. But then there are some like, "Look, I just want to get to here. I'm kind of either all in. I need to quickly develop things." And I think cloud run for example is a great example of that. We were talking about announcements. I think one cool thing that we did was we combined Cloud Run and cloud functions. Guess what we called it? Cloud Run Functions.

Savannah Peterson

>> Sophisticated on the marketing.

>> Sophisticated. But at least this is one of those times -

Savannah Peterson

>> It's all about decreasing complexity, Gary, that's what we're all about -

>> Well, at least it says what it does.

Savannah Peterson

>> I mean, at least it says what it does.

>> -

>> I mean, naming-

Savannah Peterson

>> Oh my gosh....

>> what is in your portfolio-

Savannah Peterson

>> Don't even get me started on this one.

>> Exactly.

Savannah Peterson

>> And the spelling sometimes.

>> Exactly.

>> But the thing is, so if you want to focus on a functional programming model, you can. But then if you needed to tweak something, you can. And even if you started with a functional programming model, you could technically deploy that on Cloud Run as a function. Even on GKE, because it's the same sort of portable artifact. So I think we start to see that moving back and forth. The other thing you were mentioning about ... The interesting thing in the AI conversations is the ... I guess it's not surprising, but sort of to Bobby's point of the sneakers and the jet pack, it's like we have some of the craziest ... Not craziest, but craziest requirements from customers. And so, one time you're getting in and talking about what's the right thing to do on this model, low-level networking, RDMA, infinite band, blah-blah-blah. How do I ... Whatever. Loading. And then on the other side, well, I just want to serve this. How do I ... And I know that Kubernetes is the right thing. Maybe we'll try a Vertex or whatever that's in there, but I'm already using it. How would I do it over here? But I don't want to see it, but I want to be able to have a sort of customization and configure it. So it's sort of these two extremes. And we were with another guy on our team yesterday and he's like, "Wait, who are we talking to today, Gary? I have to get in the right mindset before I start spitting out acronyms soup, like VLM, TGI." I mean, there's acronyms everywhere. But I think that's been the interesting side is that sort of balance, that super, super high-tech and then like, "Hey, we know this is the way you want it. We want simplicity. Can you just make this easier? And how to do it." So I think it's kind of ... Maybe someday we'll get to the middle.

>> I think what we're seeing is I think the products have range, Kubernetes has a lot of range, to Gary's point, but the customers have a lot of range. So you need to really-

Savannah Peterson

>> Very much so....

>> think about what is the problem that the customer's trying to solve. And there are different parts of their journey. You can have the same problem and you just want the answer. You're still trying to validate the question in some of the cases. So we got to listen, we got to leave our preconceived notions at the door and just make sure that we're listening a lot more than we're talking.

Savannah Peterson

>> Could you say that louder for all the people in the back, just in the world? Because I think we would have a much calmer planet in a lot of different ways-

>> We would....

Savannah Peterson

>> if we all just did a little more listening.

>> Just a little bit.

Savannah Peterson

>> All right, this has flown by quite frankly, and I knew that it would. I have one final question for you both. You know where I'm going with this, Bobby, so I'm going to give it to you first so Gary has a second, since he's getting used to this fun ride we put him on here on theCUBE. When we're in London at the next CubeCon, what do you hope to be able to say then that you cannot say now?

>> What do I hope to be able to say then that we can't say now? I hope to be able to say, so back to what 50% or so of the people here are here for the first time. I hope to be able to say that we gave people on-ramps who were not in the community for years already. We made room on the platform for people that are the newbies. On-ramps for the customers, the contributors, that that's what we did. And then I got to give you one Bobby-ism before we go.

Savannah Peterson

>> Please.

>> Which is, I believe everybody wants the blessing, nobody wants the assignment. We all want the benefits of Kubernetes, but what Google's trying to do is we're giving the builders the building blocks before they need them so we can take off the limits, so you can do something amazing.

Savannah Peterson

>> You always . We got to get as many Bobby-isms in every segment. I hope you bring a few more to our 3:30 later today as well.

>> Thank you.

Savannah Peterson

>> Gary, what about you? What do you hope to be able to say?

>> I'm going to go on the pure tech side just for fun. I mean, somebody should be able to use it, but I guess I'll hint towards it is, we've talked about scaling things up in terms of size. I really want to talk about scaling things up in terms of speed.

Savannah Peterson

>> Yeah, baby.

Savannah Peterson

>> I guess, I'll sort of leave it at that. But we're making a lot of investments in those area and it's not what I want to actually tell you. It's what I want to show you-

>> Nice....

>> is that.

Savannah Peterson

>> Speaking my language all around here, gentlemen.

Savannah Peterson

>> I'm a YOLO. I get up there, live demo, share something fast. So that's hopefully ... And I guess we'll be able to talk next year ... Kubernetes turned 10 this year. GKE turns 10 next year.

>> -

Savannah Peterson

>> So we'll be throwing a birthday party in London.

>> Yeah, of course.

>> -

Savannah Peterson

>> And doing a speed demo, which I very much look forward to. And I'm going to be thinking about glow-ups for the rest of the day.

>> There you go.

Savannah Peterson

>> I think mine was probably when I got my braces off.

>> That was good.

Savannah Peterson

>> I think that was the light switch for Savannah-

>> -...

Savannah Peterson

>> as an on-camera personality, versus a face-

>> Nice....

Savannah Peterson

>> that was made for radio. On that note, Bobby and Gary-

>> Good one....

Savannah Peterson

>> you are an absolutely fabulous duo. I see exactly why you wanted to bring Gary. You did not disappoint.

>> Oh, thank you.

Savannah Peterson

>> He was not over-hyping you. This was a blast. Thanks for taking the time. And congrats again on 65.

>> Thank you.

Savannah Peterson

>> Maybe we can be doing in 10 years, 100K. Next year, it's going to be a whole lot of fun for theCUBE.

Savannah Peterson

>> In 100 years, we won't talk about the number of nodes. It'll just happen.

Savannah Peterson

>> Oh, yes. It'll already be taken care of. I love that.

>> No limits.

Savannah Peterson

>> No Limits is right, baby. That's definitely a part of the attitude.

>> No limits. Love it.

Savannah Peterson

>> Speaking of, thanks for you always having no limits to my terrible jokes on the show.

Savannah Peterson

>> Oh dude, I think you're getting better. We're all good.

Savannah Peterson

>> Thanks. That's quite the compliment. And I hope all of you are putting up with my cheeky sense of humor and our fabulous guests and their amazing insights here in Salt Lake City, Utah, day two of CubeCon North America. My name's Savannah Peterson. You're watching theCUBE, the leading source for enterprise tech news.