 Live from San Francisco, it's theCUBE. Covering Google Cloud Next 19, brought to you by Google Cloud and its ecosystem partners. Welcome back to San Francisco everybody. You're watching theCUBE, the leader in live tech coverage. And we're here at the new improved Moscone Center covering Google Next 2019. I'm Dave Vellante with my co-host Stu Miniman. Stu, good to see you. Russell Warman is here. He's the head of infrastructure at AutoTrader UK. Russell, thanks for coming on theCUBE. You're very welcome. AutoTrader UK, not to be confused with the US AutoTrader. You guys are separate companies. Yep, completely separate. So we operate the largest automotive marketplace in the UK. Something like the 16th busiest website in the UK. Around 55 million cross-platform visits each month. On average, probably about 500,000 vehicles that we're advertising out to our consumers. Yeah, AutoTrader's awesome website. If you're, it's our second biggest purchase, typically behind the home. And so Stu was asking you earlier if digital transformation meant anything to you and you said, well, we kind of went through it 10 years ago. And it's kind of true. You guys are like born in the internet company, but talk about that. So our heritage is we've been around for 40 years. We started producing magazines back in 1977. And then in 96, we launched our first website and we basically just migrated those revenues from our print business online. And then about 2013, we stopped printing magazines altogether, so we're a truly digital business now. And people talk about digital transformation, oftentimes in the context of data. So maybe talk about some of the things that you're doing in your business, some of the challenges that you face, where data fits in, and then we'll really get into it. Okay, so we've got a couple of areas where we've been on premise for our data centers. And we've had that strategy for probably about 15 years. And we've started to reach limits in terms of how we manage capacity within there. And what we've found is using cloud services has really enabled us to unlock capabilities, particularly around things like data. And then more recently, when we're thinking about using Kubernetes and SDO, we've been able to take advantage of things like some of the security features, mutual TLS and service discovery. All right, so Russell, bring us inside your organization a little bit. You're head of infrastructure. What does that mean to your business today? So we basically run the platform that runs AutoTrader.co.uk. And we're responsible for making sure that's available 24-7-3-6-5. So, and we want to make sure that we're able to give our internal customers the ability to release new applications, new features as quickly as they want, but also that we're ensuring that our consumers, our retailers, get a great experience that it's fast-performant and secure. Yeah, and give us, tease out a little bit, just kind of the scope, how many people you have and how do you balance the stuff that kind of is under your purview and the stuff that you manage that is outside of your four walls? So there's about 25 people within the team. There's probably about 10 people building the infrastructure and about 15 people that are responsible for monitoring it. So we're not a big organization in terms of operations infrastructure. And we have to work very closely with our product squads, because we manage some elements of it, but we're dependent on them managing other elements as well. So we're trying to unlock platform capabilities so that we've got consistency that we're trying to provide those capabilities rather than going off and finding something else. You've got infrastructure on-prem and you've got infrastructure in the cloud, just the cloud services, right? Yeah. So oftentimes those aren't the same. No. Right, talk about your situation and you're on a journey, I presume. What's that journey like? So we started about 14 months ago building out our cloud services and probably in the last four months we've really started accelerating that migration plan of applications moving up into Google Cloud. And our ambition is to get that done in 18 months because the complexity of managing multiple environments is something that we want to try and avoid and minimize that. So is the aspiration to go all in? Absolutely. Absolutely. We've taken a view that for services that we can, we'll take managed services and then everything else will be cloud first. All right, I believe I heard you say Kubernetes is in the mix now. Can you explain where that fits, how you're using it? So that's the platform that we're using to build out all our applications in the cloud. So that's the sort of landing point. Okay, are you using any Kubernetes in your data center today? No, just all from a cloud. That's the platform layer that you use in the public cloud. Yeah, absolutely. So when we talked about should we use Kubernetes we did discuss should we do it on-prem or do it in the cloud? And we just figured that we didn't want the overhead of managing it ourselves on-prem. We thought it'd be better just to take it as a service almost and manage it up in the cloud. Most of the shows we go to, many of them anyway, Stu and I and theCUBE, they talk about multi-cloud and we often say multi-cloud is a function of multi-vendor. Certainly we heard today from Google about multi-cloud. What's your cloud strategy? Because there's another camp that says, well, if you do multi-cloud it's more complex, it's less secure, it's more expensive. Are you trying to be kind of unit cloud or is it horses for courses? So we're with three clouds today. So our data platform sits partly in GCP but then partly in another cloud provider. And then we're also for our enterprise applications we're using another cloud provider because it makes sense to do so. So we want to use the right cloud for the right applications. I think that most of our customer-facing applications will end up in GCP but some of the back-end services may end up in other cloud providers. Okay, so it's strategic fit based on the application. Absolutely. And so what is the wheelhouse of Google in your view from a customer perspective? So the fact that Kubernetes was developed by Google is a really strong play there. It gives you confidence that they know how to look after it. Things like STO have made a massive difference to our organization as well. Part of the reason why we've ended up using GCP was based on using STO and that was around solving a specific problem that we had. So those are the sort of two sort of areas that we focused on. Go ahead. I'm sorry, just when you look at, you've got that hybrid multi-cloud environment that we find is pretty much the standard today. Can you speak a little bit of the management layer, how you look at that? Is it good enough today? And what could the industry do to make it even better? It's good enough for our needs. I think that the challenge that you've always got is data ingress and egress between clouds as you want to move data or query data and making sure that you can do that in a secure way. That's probably the biggest challenge that you'd have and also around the cost of doing so. You know, those are probably the bigger challenges. And the challenge is the cost and time of moving data? Is it security or the above? Maybe you could dig into that a little bit. It's probably a little bit of all the above, if I'm honest. I think you can do so much to secure it, like private VPNs between your cloud providers. It's, I guess, it's the time and actually moving some of that data between the clouds that's a challenge. And then the cost, like I say, it's difficult to predict how much it might cost you to move some of that. What are the big challenges from a business perspective that are driving your technology strategy? Obviously, you want consumers coming to your site. You want to make that as rich as possible for them. You've got a monetization strategy as well, but talk a little bit about the business drivers that are affecting your technology. Okay, so consumer demand is constantly changing. Technology is massively disrupting how people think about and search for cars. We can see that there's a demand in the marketplace for people just to be able to choose a car and have that customized and delivered. They want to have complexity around how they look for finance on their cars. They want all that sort of taken away. They don't want to have to turn up to a car showroom and then go through the same conversation that they've with a salesperson that they've just gone through online. They want that journey to be seamless. So there's some challenges that the industry faced in trying to do that. And where we're trying to help our retailers is by providing those services that customers want. So to be able to understand what the par-tex valuation is on their car before they turn up to the dealership and know that it's going to get honored because they've got a trust in it. So they can understand what they can spend on a monthly fee and what cars are available. So we're trying to change what we provide to our consumers and also to support our retailers manage that change as well. Yeah, Russell, I think it's a really great point you make. We know many industries and especially your industry is changing really fast. So what do those stress and strain mean to the infrastructure team? Can maybe talk a little bit about the relationship of kind of business to IT and how that's changed in the last few years. So probably over the last 10 years, IT's always been seen as the sort of the blocker in terms of making sure that new products and features can become available quickly. And our role is really to make sure that we're providing the infrastructure at the right time for people so that they can basically just dial it in when they need it. So if they want to release a new application, they don't need to come and speak to somebody in ops to sort of provision a server. They just create a pipeline, deploy their application and then the server's provisioned at the same time. And we do things like checking for OWASP vulnerabilities at the same time and making sure that the application and the infrastructure is working hand in hand. So we've taken a lot of that dependency away from people sort of with handoff points and everything. You're a business that has obviously transformed over the years. I think you said you started in the 60s. 70s as a publisher, basically paper, printing presses and the like. 1996 is when you went online, is that right? So that early days pretty much of the internet sort of the heart. And so you've proven that you could transform forward thinking. I'm wondering what the conversation might be like inside the company about things like autonomous vehicles. Is there a day that comes where owning your own vehicle is the exception rather than the norm? What implications that might have for your business? It's very hard to predict obviously, but are you having those kind of conversations? You have other lines of business that you're launching to? So you're absolutely right. Those things are an area of concern and we've got people constantly looking at new and emerging changes and threats across the whole industry. And they are absolutely looking at how that might impact our business. What we're actually seeing though through research that we've done is people still aspire to own cars. So car ownership is still something that people want to look at. With regards to the second part of your question, we are automotive focused. That's where we are. We don't operate outside of the UK and Ireland. We absolutely focus on our core and making sure that the services that we offer around that really support our customers' needs, whether that be consumer or retailer. Well, it's interesting. The auto industry really hasn't been wildly disrupted. I mean, you certainly see Tesla and people talk about autonomous vehicles, but still the big car makers are still doing quite well. They dominate the market. So it's going to take some time and there's some skeptics out there, but it's going to be interesting to see how you guys evolve. What other opportunities you go after? I want to ask you, so Google Next is coming to London in November. How important is it for you as somebody who's based in the UK to have Google have local events like that? I mean, obviously you're traveling a far distance to come to this event. How important is that to you as a customer? It's important for us to be able to get more people involved in the conversation. There's obviously a cost and time effort in terms of coming out to something here, but it's, I mean, the scale of this is enormous in comparison to what we'd see in the UK, but it just means that having that presence in the UK means that we can just get more of our engineers in front of the right people to have a better conversation, understand more about what's coming and how we might be able to use that within our business. Things that, so you've just sort of, you're early on in your cloud journey, but knowing what you know to get, it's early on. Things you might do differently if you had a mulligan you want to do over again, or advice you'd give to your colleagues and peers? So I think the big thing that we found is modernizing applications before you start moving them into the cloud. I think there is a tendency probably, and we were in the cloud probably about nine years, 10 years ago, and we bought those workloads back on-prem, and the reason why we did that is because we didn't treat them differently. We managed those instances in the same way that we managed on-prem, which is completely the wrong way to do it, in my opinion. So we needed to change our mindset in terms of how we managed the infrastructure. You need to make sure that you re-architect your applications that you are taking advantage of the features and functionalities, things like autoscaling, that you plan for failure, all those things that you typically control on-prem, you have to think about differently in the cloud. We talk about there's a lot to do of changing the operating model, right? The long pole in the tent is, modernizing those applications, any change, migrations is really tough, so do you have a timeframe? How long will it take you to get to that all-in that you're planning on? 18 months is our target. That's our target. And not just a lift and shift. No, no, no, so that is a modernization. But the approach that we've taken and is to try and give the people the confidence and the courage really to go at it is we've picked off some easier applications, so this journey about being cloud native, we started six or seven years ago and we've started getting people building their applications on our private cloud and those ones have been really easy to move on to the public cloud. And some of the applications that have been a little bit trickier that have been on our sort of traditional virtualization platform, we're trying to pick those off in parallel, so you do one that's really easier, one that's a bit harder, so you give people confidence and then you're trying to solve some of the tricky problems alongside it and I think that just incrementally gives you more confidence that you're moving in the right direction. It's a great strategy. It's like skiing on the blacks and then going to the greens and then making your moves. Just a little bit harder each time you go at it, but yeah. Russell, thanks so much for coming to theCUBE. It was great to have you. It's a pleasure, thank you very much. Okay, thank you for watching. We'll be right back from Google Next in San Francisco.