 Good evening, brilliant humans. My name is Savannah Peterson, and very delighted to be streaming to you live from theCUBE Studios here in Motor City, Michigan. I've got John Furrier on my left. John, this is our last interview of the day. Energy just seems to keep oozing. How you doing? How you doing? Day two, three days of coverage of theCUBE. Love the segment. This one's great because we have a practitioner who's implementing all the hardcore talks. It's been awesome. Can't wait to get into it. Yeah, I'm very excited for this one. If it's not very clear, we are community focused. Community is a huge theme here at the show at KubeCon, and our next guests are actually a provider and a customer. Turning it over to you, Lisa and Jake, welcome to the show. Thank you so much. It's great to be here. It is our pleasure. Lisa, you're with Cockroach. Just in case the audience isn't familiar, give us a quick little sound bite. We're a distributed SQL database, highly scalable, reliable, the database you can't kill, right? We will survive the apocalypse. So very resilient. Our customers mostly retail, fintech, gaming online gambling, they need that resiliency, they need that scalability. So the indestructible database is the elevator pitch. And the success has been very well documented. Valuation obviously is a scorecard, but huge customers. We were at the Escape 19 just for the record. The first ever multi-cloud conference hasn't come back. Baby, it'll come back soon. Yeah, well we did a similar version of it just a month ago. And that was before Cockroach. I was a different company there, talking a lot about multi-cloud. But I've been a conference a couple of years now and I run community, I run developer relations. I'm still also a CNCF ambassador. So I lead community as well. I still run a really large user group in the San Francisco Bay Area. So we've just been in community for years. So take us through the use case. Jake's story, set us up. Well I would like Jake to take him through the use case and Cockroach is a part of it. But what they've built is amazing and also Jake's history is amazing. So you can start wherever you like. Yeah, sure. I'm Jake. I'm CEO and co-founder of AuthZed. AuthZed is the commercial entity behind SpiceDB. And SpiceDB is a permission service. So a permission service is something that lets developers and lets platform teams really unlock the full potential of their applications. So a lot of people get stuck on, my RBAC isn't flexible enough. How do I do these fine-grain things? How do I do these complex sharing workflows that my product manager thinks is so important? And so our service enables those platform teams and developers to do those kinds of things. What's your infrastructure lead? What's your setup look like? Are you guys looking like on the back end? Sure, yeah. So we're obviously built on top of Kubernetes as well. One of the reasons that we're here. So we use Kubernetes. We use Kubernetes operators to orchestrate everything. And then we use CockroachDB as our production data store, our production back-end data store. So I'm curious, because I love when these little match-makes come together. You said you've now been presenting on a little bit of a road show, which is very exciting. Lisa, how are you and the team surfacing stories like Jakes? Well, I mean, any place we can. Obviously all the social medias, all the blogs. How are you finding them? How did you? Like from our customers? We have an open-source version. So people start to use us a long time before we even sometimes know about them. And then they'll come to us and they'll be like, I love Cockroach and like, tell me about it. Like, tell me what you built. And if it's interesting, you know, we'll try to give it some light. And it's always interesting to me what people do with it. Because it's an interesting technology. Like what they've done with it. I mean, the fact that it's globally distributed, right? That was like a really important thing to you. Totally, yeah. We're also long-term fans of Cockroach. So we actually all work together out of Workbench, which was a co-working space and investor in New York City. So we go way back. We knew the founders. I'm constantly saying like, if I could have invested early in Cockroach, that would have been the easiest check I could have ever signed. Yeah, that's awesome. And we've been following that too. And you guys have been now using them. For folks that are out there looking to have the same challenges, what are the big challenges on selecting the database? I mean, as you know the history of Cockroach and your originator story, folks out there might not know. And they're also going to choose a database. What's the big challenge that they can solve that kind of comes together? What would you describe that? Sure. So we're, as I said, we're a permission service. And the data that you store in a permission service is incredibly sensitive. You need it to be around, right? You need it to be available. If the permission service goes down, almost everything else goes down because it's all calling into the permission service. Is this user allowed to do this? Are they allowed to do that? And if we can't answer those questions, then our customer is down, right? So when we're looking at a database, we're looking for reliability. We're looking for durability, disaster recovery. And then permission services are one of the only services that you usually don't shard geographically. So if you look at like AWS's IAM, that's a global service, even though the individual things that they run are actually sharded by region. So we also needed a globally distributed database with all of those other properties. So that's what led us to Cockroach. This is a huge topic, Savannah. We've been talking about it all week. The cloud is essentially distributed database at this point. It's a distributed system. So distributed database is a hot topic and not really well reported. A lot of people are talking about it, but how would you describe this distributed trend that's going on? What are the key reasons that are driving it? What's making this more important than ever in your mind, in your opinion? I mean, for our use case, it was just a hard requirement, right? We had to be able to have this global service. But I think just for general use cases, a distributed database has that shared nothing architecture that allows you to kind of keep it running and horizontally scale it. And as your requirements and as your applications needs change, you can just keep adding on capacity and keep adding on reliability and availability. I'd love to get both of your opinion. You've been talking about the phases of customers. The advanced, got Kubernetes going crazy, distributed, super alpha geeks. And you got the people who are building now and then you got the laggers who are coming online. Where do you guys see the market now in terms of, I know that alphas are all building all the great stuff and you guys had great success with all the top logos. They're all doing hardcore stuff. As the mainstream enterprise comes in, where's their psychology? What's on their mind? Do you share any insight into your perspective on that? Because we're seeing a lot more of IT folks becoming like real cloud players. Yeah, I feel like the mainstream enterprise hasn't been lagging as much as people think. You know, it's certainly there's been pockets in big enterprises that have been looking at this. And as distributed SQL, it gives you that scalability that is absolutely essential for big enterprises. But also it gives you the multi-region, you know, you have to be globally distributed. And for us, for enterprises, you know, you need your data near where the users are. I know this is hugely important to you as well. So you have to be able to have a multi-region functionality. And that's one thing that distributed SQL lets you build and that what we built into our product. And I know that's one of the things you like too. Yeah, well, we're a brand new product. I mean, we only founded the company two years ago, but we're actually getting inbound interest from big enterprises because we solve the kinds of challenges that they have. And whether, I mean, most of them already do have a cockroach footprint, but whether they did or didn't, once they need to bring in our product, they're going to be adopting cockroach transitively anyway. So you're built on top of cockroach, right? And SpiceDB, is that open source? It is, yep. Okay, and explain the role of open source in your business model. Can you take a minute to talk about the relevance of that? Yeah, open source is key. My background is before this, I was at Red Hat. Before that, we were at CoreOS, so CoreOS acquisition. And before that, we made- One of the best acquisitions that ever happened for the value, that was a great team. Yep, we had fun. And before that, we built Quay. So my co-founders and I, we built Quay, which is a first private Docker registry. So CoreOS and all of those things are all open source, are deeply open source. So it's just in our DNA. We also see it as part of our go-to-market motion. So if you are a database, a lot of people won't even consider what you're doing without being open source. Because they say, I don't want to take, I don't want to end up in an oracle situation again. Yeah, oracle meaning they gouge you, get you locked in, get you in a headlock, increase prices. Yeah. I can get out late. I got triggered. Real fast. I was going to say, do we need to talk about your PTSD here? Or what's going on? Or in the black box, I mean, we have 20,000 stars on GitHub because we've been open and transparent from the beginning. Well, and both of your projects were started based on Google Papers, right? That is true, yep, and that's actually, so we're based off of the Google Zanzibar paper, and as you know, Cockroach is based off of the Google Spanner paper. And in the Zanzibar paper, they have this globally distributed database that they're built on top of. And so when I said, we're going to go and we're going to make a company around the Zanzibar paper, people would go, well, what are you going to do for Spanner? I was like, easy, Cockroach, they've got us covered. I know the guys and my friends. Yeah. So the question is, why didn't you get into the first round of Cockroach? She said, I'll tell you something about that. The question he did answer, though, was one of those age-old arguments in our community about pronunciation. We used to argue about Kuei. I always called it Kee, of course. And the co-founder obviously knows how it's pronounced. You know, it's the XCD argument. It's the Kube cuddle versus Kube control versus Kube CTL. Kuei from the co-founder, that is end of argument. You heard it here first. And we're keeping it going with Ozzed. So, awesome. A lot of people will say Ozzed or, you know, so we just like to have a little ambiguity. You got to have some semantic arguments, arm wrestling here. I mean, it keeps everyone entertained, especially over the weekend. What's next? You got obviously Kubernetes in there. Can you explain the relationship between Kubernetes, how you're handling SpiceDB? When does the Kubernetes piece fit in and where is that going to be going? Yeah, great question. Our flagship product right now is Ozzed Dedicated. And in Ozzed Dedicated, what we're doing is we're spinning up a single tenant Kubernetes cluster. We're installing all of our operator suite, and then we're installing the application and running it in a single tenant fashion for our customers in the same region, in the same data center where they're running their applications to minimize latency. Because this is an authorization service, latency gets passed on directly to the end user. So everybody's trying to squeeze the latency down as far as they can. And our strategy is to just run these single tenant stacks for people with the minimal latency that we can and give them a VPC dedicated link. Very similar to what Cockroach does in their dedicated product. And the distributed architecture makes that possible because it's lighter weight, not as heavy. Is that one of the reasons? Yep, and Kubernetes really gives us sort of like a level playing field where we can say we're going to take the provider, the cloud providers Kubernetes is offering, normalize it, lay down our operators and then use that as the base for delivering our application. Jake, you made me think of something that I wanted to bring up with other guests but now since you're here, you're an expert, I want to bring that up. The talk about super cloud, it was we coined that term, but it's kind of multi-cloud, is that having workloads on multiple clouds is hard. I mean, they are workloads on clouds. But the complexity of one cloud, let's take AWS, they got availability zones, they got regions, you got now data issues in each one. Being global is not that easy on one cloud. Nevermind all clouds. Can you share your thoughts on how you see that progressing because when you start getting this distributed database, a lot of good things might come up that could fit into solving the complexity of global workloads. Can you share your thoughts on scoping that problem space of geography? Can you mention latency, that's huge. What are some of the other challenges that other people have with global? Yeah, absolutely. When you have a service like ours where the data is small but very critical, you can get a vendor like Cockroach to step in and to fill that gap and to give you that globally distributed database that you can call into and retrieve the data. I think the trickier issues come up when you have larger data. You have huge binary blobs. So back when we were doing Quay, we wanted to be a global service as well, but we had terabytes, petabytes of data that we were like, how do we get this replicated everywhere and not go broke? So I think those are kind of the interesting issues moving forward is what do you do with those huge data lakes, the huge amount of data. But for the smaller bits, like the things that we can keep in a relational database, we're happy that that's quickly becoming a solved problem. And by the way, that data problem also is compounded when the architecture goes to the edge. Totally. I mean, this is a big issue. Exactly, yeah. Edge is something that we're thinking a lot about too. We're lucky that right now the applications that are consuming us are in a data center already. But as they start to move to the edge, we're going to have to move to the edge with them. And it's a story that we're going to have to figure out. All right, so you're a customer of Cockroach. What's the testimonial? If I put you on the spot and say, hey, I'd like to work with these guys, you know, what's the, what's the, you know, the founders. So, you know, it's a good description, little advice, but we'll hold you on it. Yeah, working with Cockroach has been great. We've had a couple of things that we've run into along the way. And we've gotten great support from our account managers. They've brought in the right technical expertise when we need it. Because what we're doing with Cockroach is not, you couldn't do it on Postgres, right? So it's not just a simple rip and replace for us. We're using all of the features of Cockroach, right? We're doing, as of system time queries, we're doing global replication. We're, you know, we're consuming it all. And so we do need help from them sometimes and they've been great. And that's natural as they grow their servers. I mean, the world's changing. Well, I think one of the important points that you mentioned with multi-cloud, we want you to have the choice. You know, you can run it in clouds, you can run it hybrid, you can run it on-prem, you can do whatever you want. And it's just, it's one application that you can run in these different data centers. And so, Willie, it's up to you. How do you want to build your infrastructure? And one of the things we've been talking with the super-cloud concept that we've been interested in getting a lot of contrary, but people are leaning into it, is that it's the refactoring and taking advantage of the services, like what you mentioned about Cockroach. People are doing that now on cloud, going, the lift and shift market's kind of had its time. Now it's like, hey, I can start taking advantage of these higher level services or keep it going, of someone else's stack and refactoring it. So I think that's a dynamic that I'm seeing a lot more of, and it sounds like it's working out great in this situation. I just came from a talk and I asked them, you know, what don't you want to put in the cloud and what don't you want to run in Kubernetes or on containers? And the, yeah, and the customers that I was on stage with, one of the guys made a joke and he said, I would put my dog in a container. He was like in the category of like, I'll put everything in containers. And these are, you know, including like, mission critical apps, heritage apps, and they don't want to see legacy anymore, heritage apps. These are huge enterprises and they want to put everything in the cloud, everything. You just don't want your dog to get stuck on the airplane when it's on the tarmac. Oh my God. Oh, don't take that analogy literally. Don't think about that. Let's not containerize our pets. If we can't afford it, let's not. So, I mean, going macro and especially given where we are, CNCF, it's all about open source. Do y'all think that open source builds a better future? Yeah, and a better past. I mean, this is so much of this software is founded on open source. We wouldn't be here really. I've been open source community for many, many years. So I wouldn't say I'm biased. I would say this is how we build software. I came from like in high school, we're all like, oh, let's build a really cool application. Oh, you know what, I built this because I needed it, but maybe somebody else needs it too. And you put it out there and that is the ethos of Silicon Valley, right? That's where we grew up. So I've always had that mindset and social coding and why have three people working on the same thing when one person we could share. It's so inefficient, all of that, yeah. So I think it's great that people work on what they're really good at. Now you need some standardization, you need some kind of control around this whole thing, sometimes some foundations to herd the cats. But it's great, which is why I'm a CNCF ambassador and I spend a lot of time in my free time talking about open source. Yeah, it's clear how passionate you are about it, Jake. This is my second company that we founded now and I don't think either of them could have existed without the base of open source, right? Like when you look at, I have this cool idea for an app or a company and I want to go try it out. The last thing I want to do is go and negotiate with a vendor to get like the core data component to even be able to get to the prototypes. And pay too, by the way. Yeah, or hire a bunch of PhDs to go and build that core component for me. So yeah, I mean, nobody can argue. It truly is, I got to say, a best time. If you're a developer right now, it's awesome to be a developer right now. And it's only going to get better as we were ripping in the last session about productivity. We believe that if you follow the digital transformation to its conclusion, developers and IT aren't a department serving the business. They are the business and that means they're running the show, which means that now their entire workflow is going to change. It's going to have to be leveraging services, partnering. So, yeah, open source just fills that. So the more code coming up, it's just no doubt in our mind that that's happening and will accelerate. So, you know. No one company is going to be able to compete with a community. 50,000 users contributing versus you riding it yourself in your garage with your dog. It's people driven too. It's humans. It's humans working together. And here you'll see, I won't say horse training is a bad term, but like as projects start to get traction, hey, why don't we come together as the world starts to settle and the projects have traction, you start to see visibility into use cases, functionality. Some projects might not be, they have to kind of, you're going to see more, kind of core lesson. Then not every feature is going to be developed over so. I mean, you know, this is why you connect with truly brilliant people who can architect and distribute SQL database. Like, who thought of that? It's amazing. It's as our friend used to say. So let me ask you a question before we wrap up by the time. What is the secret of Kubernetes success? What made Kubernetes specifically successful? Was it timing? Was it the unambitious nature of it? The unification of it? Was it, what was the reason? Why is Kubernetes successful, right? And why nothing else? Well, you know what I'm going to say. So I'm going to let Jake answer first. Don't let Jake, you go first. Oh boy. If we look at what was happening when Kubernetes first came out, it was Mesosphere, was kind of like the big player in the space. I think Kubernetes really, it had the backing from the right companies. It had the, you know, had the credibility. It was sort of loosely based on Borg, but with the story of like, we fixed everything that was broken in Borg and it's better now. Yeah, so I think it was just kind, and obviously people were looking for a solution to this problem as they were going through their containerization journey. And yeah, I think it was just right place. The timing consensus of, hey, if we just let this happen, something good might come together for everybody. That's what I felt. I think it was right place, right time, right solution. And then it just kind of exploded. When we were at CoreOS, Alex Pulvia, our CEO, he heard about Kubernetes and he was like, you know, we had a thing called FleetD. We had a tool called Fleet and he's like, nope, we're all in on Kubernetes now. And that was an amazing, amazing decision. Yeah. It's clear we can feel the shift. It's something that's come up a lot this week is the commitment, everybody's all in, people are ready for their transformation and Kubernetes is definitely going to be the orchestrator that we're leveraging. And it's an amazing community, but it was, we got lucky that the, the foundational technology, I mean, you know, coming out of Google based on Go, completion based on Go, it's no coincidence that this sort of nature of, you know, pods horizontally scalable, it's all fits together. I mean, those make sense. Yeah, I mean, no offense to Python and some of the other technologies that were built in other languages, but Go is an awesome language that's so, so innovative, innovative things you could do with it. Oh, definitely. Jake, I'm very curious since we learned on the way in, you're a Detroit native. I am, yep. I grew up in the, in Warren, which is just a suburb right outside of Detroit. So what does it mean to you as a Michigan born bloke to be here, see your entire community invade? It is, I grew up coming to the Detroit auto show in this very room. That brought me to Detroit the first time, love NAIS been there with our friends at Ford just behind us. And it's just so interesting to me to see the accumulation, the accumulation of tech coming to Detroit because it's really not something that historically has been a huge presence. And I just love it. I love to see the activity out on the streets. I love to see all the restaurants and coffee shops full of people. I'm just, I might tear up. Yeah. I was wondering if it would give you a little bit of that hometown pride and also the joy of bringing your community together. I mean, this is merging your two probably most core communities. Yeah. Your youth and your career. It doesn't get more personal than that really. It's just been, it's been really exciting to see the energy. Well, thanks for going on theCUBE. Thanks for sharing. Appreciate it. Thanks for having us. Yeah, thank you both so much, Lisa. You were enjoying a ball of energy right when you walked up. Jake, what a compelling story. I really appreciate you sharing it with us. John, thanks for the banter and the fabulous questions. I'm glad I could help out. Yeah. You do a lot more than help out, sweetheart. And to all of you watching theCUBE today, thank you so much for joining us. Live from the Troit, the CUBE Studios, my name is Savannah Peterson, and we'll see you for our event wrap up next.