 Live from San Francisco. It's theCUBE. Covering Google Cloud Next 2018. Brought to you by Google Cloud and its ecosystem partners. Hello everyone, welcome back to theCUBE coverage live in San Francisco at Moscone South for Google Next. Google Cloud Next 18, I'm John Furrier, Dave Vellante, next guest is Jennifer Lin, Director of Product Management at Google Cloud. Featured in Forbes as one of the power women at Google Cloud. Congratulations on your Forbes distinction. Thanks so much John. Great to see you. So we had a chat before the event a couple weeks ago, leading up to it around Istio, Kubernetes. You're in charge of a lot of the cool, I would say the modern middleware that's going on. I want to say middleware kind of in quotes, it's not really middleware, it's cloud, it's horizontally scalable. Take a minute to explain some of the areas you're working on and then the importance of Istio's announcement to the 1.0 generally available. Huge news, it's kind of nuance, it's not as big as the cloud services platform and some of the Cisco relationships, but huge progress in this services, microservices. This is the key part, take a minute to explain. Yeah, we're really excited to get to this week and I think the announcement of cloud services platform of which obviously the evolution of Kubernetes and Istio are a key part. Now we've kind of changed the way people manage container environments and now people are really writing, really innovative services and microservices and the ability to manage that easily is really what Istio is all about and do that in a secure way. So we had the CEO Sundar Vakai on stage, Google proper as Diane Green was also on stage. Sundar made a comment, this is the only event I could make like containers and Kubernetes joke. Yes, with the big containers. Translate, he's smart, he knows tech, very strong tech culture here, but I want to, Jennifer, explain to the people why Google cloud is differentiating around APIs and services and open source. Why is that so important? Yeah, I mean at heart I think we really are a software innovation company and Google's a company of developers that want to do creative things with software. As Diane said this morning, I think the sort of ability to do that in a way that hides the complexity but also excites emerging developers with all the things that they can do. I think that's what we're seeing in cloud and originally we started very much with the cloud natives who were doing very new types of applications for consumer applications. As Sundar said, when we moved in to kind of doing business applications and more and more people were developing enterprise applications with sort of a cloud native model, we started to see a big uptake in sort of adoption of our cloud platform and I think with a lot of the things that we're doing in security and the ability to enable administrators to kind of manage that in a more automated way, that's a lot of what I think we're differentiating around. So one of the headlines that I can see happening on either SiliconANGLE or TechCrunch or some of the blogs and publications out there is Google doubles down on Kubernetes. And the announcement of Istio's general availability of 1.0 certainly is good news. What does that mean? Why is that, and what should people know about that importance of Kubernetes doubling down as a momentum point for Google and the importance of Istio? What is the real benefit to the customer? Yeah, so we've had a managed Kubernetes environment on GCP for four years now, but before that, Orr's talked about a decade worth of understanding how to scale Kubernetes in an operational environment. So we've learned a lot of domain knowledge there that we're kind of baking into the software platform itself. And Istio sort of really models the way we do microservice management as we launch billions of containers a week. So how do we essentially secure the service environment? How do we give really good visibility? We showed the service graph where we can see the latency between two services and really hide a lot of the backend complexity that really from an operational perspective is causing a lot of toil for application developers as well as operators. I noticed Toil's a word that is being kicked around the Google community a lot. Toil being headaches, pains. But I want you to take a minute to explain for the folks that are learning about Kubernetes for the first time. Kubernetes was donated or donated by an open source but by Google. But prior to Kubernetes, you guys have been running Borg which is the internal system that has been the foundation of the scale of the service management for all of Google. Explain that important history there and how you're making Kubernetes easy to consume because most companies aren't Google. Explain the little history and then how it translates to consumption. Yeah, so I think Borg was really built and designed to keep developer agility up and make sure that developers could be very productive but we could run essentially at global scale the container orchestration environment. When Kubernetes was donated to the open source community there were some things that needed to be defined such that the abstractions could be very clean outside of a Google environment. But that framework obviously held up very well and hence the growth with Kubernetes. Istio I think similarly, it models a lot of the way that we've done service management with the service mesh within Google. Obviously the names are slightly different but there's a lot of operational domain knowledge on best practices and how to essentially enable automation at a much more granular level of applications where it's not a bunch of proprietary applications but you have a lot of loosely coupled systems coming together. So Jennifer, the maturity curve of the developer community obviously is in some bell shape. How do you, how does Google approach engaging with those developers? Are you trying to get sort of leading edge guys that want to develop software the way Google develops software? Obviously you're trying to reach a bigger market so how do you balance those two? Well and I think that's where open source is the most exciting because whether it's kids in school or very experienced developers, number one the transparency, things move so fast. A lot of that is about developer reach but also about the participation of developers to give back to the community and help evolve the system. For something like Kubernetes obviously and NSDO Google sort of bootstrapped that and donated to the community but since then we've seen just incredible participation at things like KubeCon and developer hackathons, et cetera. So that's a lot of both a model for growing the community but also just to educate and share essentially a lot of the best practices in a different type of way than most software companies I think. Well and you've worked in a lot of very successful enterprise companies, some very profitable enterprise companies. I mean I get the sense that profit is an outcome of doing good work at Google. It's like you don't wake up in the morning and say okay how am I going to make money? Say how am I going to do work and you don't seem to be stressing, I guess it helps and you have $100 billion in the balance sheet but is that the right way to think about how you guys think about the marketplace? Yeah I think the goal is very clear for us and Sundar talked about it a lot, the sort of alignment between our original mission at Google and kind of the opportunity we see in cloud. I mean data's exploding, new applications are being written in a way that really brings together worlds that didn't sort of come together before. Healthcare applications where you need to share a lot of data, people need to do research and you need to make it very easy to share but at the same time it needs to be highly secure. We're under the same pressures as any other enterprise in terms of regulatory environments, et cetera. So making all of that easy I think is the reason why open source and open ecosystems make a lot of sense to us. It's just the only way to move fast and actually make sure that we're bringing the whole community with us. Right, but not everybody takes that philosophy obviously. That's one that you're attuned to. But when you think about Google's posture in this community, I mean you start kind of late to the enterprise game, don't seem to be too stressed about it. You're developing the ecosystem. We've seen in this world some of the companies that we work with, it's Winner Take All. Is cloud just so big that there's plenty of room for everybody or is it Winner Take All in different segments? How do you think about that? I think our leadership to Diane basically really sees this as we're playing the long game and it is about driving adoption more so than essentially quarter to quarter revenue and when we're sort of reinventing how software is designed and delivered and published et cetera and shared, I think it is not going to be sort of the monetization per quarter which many of the companies I think have to be under the pressure for. Within Google, I think we really do see this as sort of the future of software and that's going to take some time. But yeah, for things that I think we see, Orr's talked a lot about, spend has gone up in many enterprise environments despite the fact that they are changing their environment. Automation is a way to bring down a lot of the cost. So we believe there's a lot of value to be captured there but we're not in a race to essentially monetize every piece of, every product we put out there. So how do you measure your success? Is it just a feeling that yeah, we're doing good work or adoption? Adoption and the happiness of our customers and the lead partners that we work with. So our leadership is very focused on that. We want to prove it out with some trusted partners and customers and I think some of those were on stage today. Make sure that it's replicable and make sure that we leave our options open because you never know what's going to happen in the next year. I got to ask you about the on-prem solution they demoed today. Obviously they kind of put a little Easter egg in the demo and then came back and said, oh, by the way, that node was on-premise and cloud. So one of the things we talked about and you've been harping on this about Kubernetes orchestrating an abstraction at higher levels of services. Both in the cloud and on-premise. It's happening now that looks like that was really elegant. Is that a demo? Is that actually shipping code? How far along are you? Where is the headroom in this? Explain this important phenomenon because this is multi-cloud and I've been really negative on multi-cloud until we see things like this. This is easy to understand. Yeah, I think now that you really have sort of workload portability and a common abstraction layer and a single point of administrative control, there's a lot you can do there. And that was really hard to do, I think, with the proprietary systems. So that wasn't just sort of a demo. I mean, a lot of customers are starting to see that they have to think about hybrid and multi-cloud in a different way. And using some of these innovative technologies with containerization, you don't have to worry about the kernel version and the OS and a lot of the toil that was in the system before. So yeah, I think we're coming at hybrid cloud and multi-cloud in a way that no other cloud provider is. And that was, I think, the start of what a lot of customers have waited for. Yeah, and certainly this is the benefit of a Kubernetes. And then Istio now has got some capabilities into a policy and that's still going to evolve. The question I want to put to you, and I think this is, I think the, I'll put the devil's advocate role here. Play the devil's other role. Shouldn't the multi-cloud be an independent group? Or if I'm going to say, okay, Google, I'm nervous. You're going to do all this stuff. There's a trust there. How do you guys answer that? The naysayers who might say, it should be an independent organization handling multi-cloud. What's the answer to that? Well, I think that's why a lot of the partners that we worked on initially was something like Istio, you know, IBM and Lyft. I mean, they also didn't want to be sort of, you know, locked into any one cloud provider. And, you know, we've done some things even in the marketplace where we believe that the future is hybrid and multi-cloud. So yeah, I think from a technology perspective, just making sure that essentially we can define those interfaces in a way that's not tied to a vendor implementation, be transparent. We have, you know, in Istio, things like Partner Mixer adapters, that ecosystem is growing very quickly. So that pluggable adapter model allows the whole ecosystem to participate. And the role of open source on all this. Obviously, Istio, we were at the Linux Foundation's CNCF covering this pretty heavily in Denmark, just passed recently when we've spoken about it. How does all the action happening here at Google Next impact open source? What's going upstream? What are some of the updates? Can you share what's going on in open source with Google? Yeah, I mean, you know, Istio 1.0 is essentially an announcement about the open source effort. I think we also saw that many of our enterprise customers want a managed environment. So just like Kubernetes, you know, we have the open source Kubernetes, which is rocking and rolling along. We have our managed Kubernetes commercial offer. Now that there's a level of maturity in the managed Kubernetes environment and people are excited that Istio 1.0 is getting sort of more mature. They want that to be a part of the evolution of their managed Kubernetes environment, which is why we're starting to see just the whole stack evolve. You know, first we've abstracted the infrastructure, now we can manage services, and then we can bring in a whole new type of ecosystem. So it's pretty exciting. So here's a philosophical question for you. Dave and I always like to talk about like old new way. So old IT is like horse and carriage and buggy and cloud is like the first car. Now you got sports car. How do you explain all the under the hood examples of the engine? Yeah. I mean, the car just drives. You don't have to feed the horses and hey, what's the new benefits that the old world won't see with cloud? Can you try to tease out and from your perspective, what are some of those things that kind of go away and say, wow, we used to do that? What are some of the things? I think, you know, even within how we build our products, we're very focused on user experience. And sometimes the user is a developer, sometimes the user is a administrator and sometimes the user is the end user. And in our case, maybe the customer's customer. So we do a lot of UX research, but like you said, there's a lot of complexity in a car, but when I drive a car, I just want to drive the car. So the user experience for the driver is very different from the mechanic who's fixing the engine. And I think, you know, there's no doubt that there is a lot of complexity in these large scale global distributed systems, but many of our enterprise customers don't want to know every little bit of how it's built. What they want to know is some declarative end state of what they want to get to the functions that they want or, you know, the application that they're trying to drive. And so I think that is sort of the maturity level that we're at, where Istio hides a lot of that complexity, provides a common service abstraction, but still gives essentially the administrators the things that they need out of the system. Well, and it speaks to, as well, and you guys talked about this in your interview, how software is being developed and how that's changing. I mean, when I deal with Spotify, if I have a problem, I don't call it like a billing department or the customer service department, I just do it. Yes. And that's the way software is going to be developed in the future, you know, versus way most enterprise. And you talk about a great customer of GCP Spotify. I use them every day as well. But yeah, that is a lot about user experience, but what they've done with machine learning to basically serve up the song that I want to hear that day based on the playlist I had before. I mean, it really is changing how software is done. So if you look at some of these old metaphors like horses versus cars, you mentioned that jobs get automated away with that old model, but yet there's new jobs are created. So I want you to talk about where, what's going away and what's evolving. Cause the value is shifting up the stack with higher level sets of services and new abstractions which are, you don't need to know all the details. Just magic happens for the customer. There's new value being created. So you can almost kind of look at the market and say, hmm, IT operations, decimated. Manual configuration, decimated. Well, I mean, that's the history of technology. I think the history of technology is sort of moving forward and automating things. You know, for Google obviously, we don't think of the software layer as just the infrastructure layer. You know, a lot of what we're trying to get to is essentially, you know, with things like machine learning and analytics. I mean, that's real business value that people really had, you know, too much toil to essentially stitch the systems together. Now as the platform evolves, I think it just becomes one stack and we can, you know, put those tools into. Is there an API administrator? I mean, cause you start to see people starting wiring services together between building blocks. It's almost the cloud model, right? So is that a API administrator? Is that like a, is it code? It's still a human component. We agree. But what is that new role? And I think, yeah, we've always had the notion of API management, you know, with cloud endpoints and our Apigee acquisition. You know, APIs are evolving with microservices and a lot of the partners that essentially, you know, have been in that space are all rebasing on something like, you know, Istio where they can do service management at a higher level. You know, the API is part of it. You know, within Google, we use things like protobufs where you have structured data and message protocols that essentially are not just an API. So I think, you know, we think about API and service management hand in hand. And you know, both of those things I think are changing. So my final question for you, I want to get your kind of advice to the anyone practitioners out there or customers that really want to take cloud native because with containers, Kubernetes and Istio, you can actually manage lifecycle of old stuff and still bring in the new. You guys do API service management, you get cloud endpoints, billing commerce, marketplace, Kubernetes serverless and Istio is kind of a focus group. What's your advice and what's coming next that people should be aware of for the folks who want to go cloud native, want to put the more gas, less break, put the pedal to the metal with cloud native and not foreclose or have to do a rip and replace, manage their existing lifecycle applications and to bring in the new with cloud native. What's your advice? Yeah, I mean, I think build for the future, make sure you don't get stuck sort of, you know, in a silo, but you know, we often see that different pace of customers and the way they're moving to cloud native. You know, our tagline for this conference was also we're bringing the cloud to our enterprise customers. They can move at their own pace. We recognize that sometimes the migration challenges are pretty tough with their legacy systems, but you know, they have a very clear, I think much clearer view now in terms of where software is going. So depending on essentially the steps they want to take, we want to enable that either natively with what we're doing with CSP or you know, enabling partners to kind of take phased approach to that end state. Awesome and ultimately the developers for the applications will win on this. Jennifer Lin, Director of Product Management at Google Cloud, here inside the queue breaking down all the action around APIs, service management, why it's important as the modern middleware within cloud enabling developers. I'm John Furrier with Dave Vellante back with more live coverage here in San Francisco after this short break. Stay with us.