 George Gilbert from Wikibon the Cube, we're at the San Jose Convention Center at the at scale conference sponsored by Facebook I'm with Melody McFessel from Google Melody, so let's talk about Google's cloud and how it's Trying to make life easier for mainstream developers What's the process of getting them on board in a in a Least disruptive way great question, so I think it really you can look at what's happening in the cloud space and see With Compute Engine, there are some companies that want the full level of control, right? We want to tweak VMs they want complete administration from GKE Google's container engine, which is getting containers out there to ease a development process and manage your service more effectively And all the way to Google App Engine right which is you focus on your code and let Google take care of all the Operations and back-end services for you and we can scale that so in terms of coming on board I think the first place that we start is we want to understand what the customers are trying to do Enterprise customers have some unique requirements. They care about security. They care about access. They care about auditing They care about logging we want to work with them to figure out What are the right components that they need to use to run their business? They also have another challenge of they have a significant investment in on-premise technology and they're trying to look at the landscape of Their system development and their internal tools and figure out how are we going to migrate? How are we going to move certain services that we have on-prem that we understand and that we control onto the cloud So we can get the cost and the performance benefit. All right, so let's let's key in on that as next question, which is how should Corporate IT prioritize Which projects to move or should they not move and just do Greenfield? That's a great question. I'm sure if you asked anyone here at at scale, they would give you a different answer I was just working with a large client last week And it was so interesting to talk to them because they really want to focus on what they're good at their core strength and I was talking to one of their technology leaders and he's saying you know I'm I'm spending so much time on this administration a part of the physical infrastructure I don't this is not my core competency And I can't keep up with the security requirements to keep this thing running So I think really evaluating What is the core thing that you're good at and look for all of the places with laser focus that you don't need to Be spending your time on so operations overhead The physical infrastructure and then I think depending on the technology organization within Enterprise how much how much interaction do you want to have around managing and running the service? Okay, so let's take let's take the example of where you have an application that you want to build on on Google App Engine And it's got a relatively homogeneous set of services How does How much of the operation and in management of The life cycle of that application can Google offload That's a great question. So we're working on many different aspects of that So the place that we like to start as a as a developer is what's the language that you're going to be working in, right? We have a way for you to get started in that particular language on our platform and And then you can sort of pick and choose the components that are going to work for you So that's kind of the the way that we're starting We're also trying to build great tools, which is the area I really care about are the developers that our enterprise customers happy? Are they able to move quickly and have a high velocity in terms of getting features out to their user communities? Having great tools in place to be able to support that one thing that I would love to bring up here is that You know developers spend over 50% of their time trying to debug and diagnose what's gone wrong with their service And we're trying to add a bunch of tools to give Developers great insight and reclaim all that time that they're having to spend through tools like cloud trace and cloud debugger So if your service is running on the cloud, you know pretty easily what's going wrong And you can go make that fixed directly in your code and push out a change. Okay, let's continue on that theme How much worse does that problem or more challenging does that problem get when the application that the Enterprise customers trying to build involves sort of multiple services whether it's polyglot database and you know sort of application services that run on Say Ruby in some part and you know Python elsewhere just a sort of a mix of things How do you manage that lifecycle? It's very complicated. We have a lot of work still to do I think this is an area Across all cloud providers that we should be investing in because I think we know that Hybrid cloud and mixed cloud using multiple cloud service providers is the way of the future And so we we have a responsibility to make that easier for enterprise customers to manage their workloads on-prem And to also manage them on different cloud services. It's not an easy thing to do right now So if you're highlighting you're highlighting some challenges we have ahead of us, but even before you get to the mix cloud There's it sounds like there's some challenges with Managing the life cycle of different technologies within a single application yes, so if you think about the process of writing code and building and and testing it well and Deploying it reliably and then keeping it up and running It is a challenge and I think there's a lot of opportunity tools in the open-source community that I see enterprise Customers adopting that if we're all contributing to it. We can make it a lot better okay, so Last point on on the mix cloud. What are the big challenges? I imagine it's like latency in in Having some services on one cloud other services on another cloud And then the whole Management model might be different so the life cycles are managed entirely differently Is that alleviated somewhat by sort of microservices where you have Some sort of abstraction between the components Yeah, again I think that there is a lot more work to do there on abstractions and setting the standards around how to interoperate these services And I think also tooling investment and tooling would be a way to help this So if you're an IT enterprise customer and you're trying to deploy on you know multiple different clouds We should have a common set and ideally an open-source tool that makes that easy for you to interact with and see what's running where our people are there people who are Moving beyond sort of a single maybe data center or a single cloud platform in terms of Helping to manage, you know mixed cloud environments. Yeah, there I think when I talked to enterprise customers They are very interested in mixed cloud environment where where they can have Choices and diversity around choices of what they run where are there any products that are beginning to Wet the appetite of enterprise IT in this area. There are some I wish I could talk about this really exciting one But in a couple months. Oh Would that be related to Google? It's a it's an open-source project that several companies are collaborating on But the intention behind it isn't it's not it's not unique in that There are we know that enterprise this is this is important for enterprise customers to be able to deploy across multiple clouds and Make it easy for them to see what's running where and so Yes, that's some tool development underway that I think and tools like that not this one specifically But tools like that that I see the community investing in that will make enterprise customers lives easier Would this be would this be just in the monitoring? sort of Segment of you know life-cycle management or would it also be in like you know the provisioning deploying change management testing There's lots of ideas in that space and if I look at you know What's it going to take to bring the next end million developers or and you know billion in revenue companies to the cloud? We have to have tooling in place like that to manage the workflow. They need it and they should expect it and and is that tooling an extension of the Development environment itself Yes, it should be it should be integrated into the development environment So that a developer who's working in a particular language on certain components of the platform has an integrated development environment that they can that they can use and be have the fastest velocity and the highest throughput it's I imagine not just for UI familiarity but because A tool like that has to capture the structure of the program as it's being written. Is that a fair way of Expressing why the sort of operational tooling has to integrate with the development tooling Well, I think it really gets to this idea of DevOps, right? Which is how you improve how you bring together development and operation, right? So that the two can have this great feedback cycle, right of I can observe my service running in the cloud I can understand what the errors are what the latency is I can Optimize it and then it feeds back into the code so this connection between running your service in the cloud and then tying back to Making faster changes in the code to push that out to their you to your users I see a lot of investment happening in DevOps tooling which really hit on Bringing development and operations together to make developers more effective. So development tools integrating with operations tools Sort of suggests that the platform players have a major role to play and that For mixed platform deployments there has to be some agreement on the standards of the operations side And I think that if tools the expectation from the community should be that if the tools work on one platform They should work on the other platform Okay, very interesting melody. Thanks This is George Gilbert. Thanks for watching You