Embracing Hybrid Cloud (Cloud Next '18)
Welcome. To Jumanji, no. You're not actually Jumanji, I was a joke, ok, alright don't. Laugh at the same time. Interrupt. Through the recording anyway, thank, you for coming I sort of triple thanks really thanks, for either, getting lunch quickly and rushing over or. Missing lunch potentially, thanks. Again for there's a key, note it seems competing with us too so appreciate you guys coming here and then of course you. Guys gonna hang out in movie theatres and super. Comfy seats super. Quiet environment, so. Hopefully we can get through the slide 6 or 7 before anyone takes it out alright there's no popcorn today unfortunately and. The. Slides are not in IMAX, I tried, but, not. An option um. Anyway, let's. Get jumping right in so my name is Matt Tavis I'm, a Solutions Architect and. What that means is I spend most of my time working with partners customers. Internal. Product teams trying, to guide people on how best to use the. Google cloud platform so, what services to consume best. Architectural, approaches. The. Essay team focuses, a lot on building the solutions, you see on the GCP website, so, the solutions section has things that are downloadable that's what we do so we spend our time we try and learn things from you and then. We try and build out templates and patterns that make sense for everyone to consume, later. On sweet R will jump up here one. Of our partners from Nutanix, and he'll, do a few slides as well we'll do a quick handoff later on but I'll let him introduce himself then, but. Without, any further ado let's jump right in so I'd love to start by talking about you know our perspective on, hybrid hybrids a pretty heavily, overloaded. Term in, the industry in, general. There's. Really no single, definition that were the captures that it's pretty broad I mean you look at the dictionary it's really. The two. Distinct. Things, that are blended together in some odd way that's, hybrid right that's so almost anything could qualify as hybrid I'm, gonna talk about some of how we feel about it and how we see Google, platform contributing, to its further development. But. First it's important to realize that hybrid is is a reality it's, not you know I've, ran, into customers in the industry have been working in the cloud for actually. Over ten years so. Always with Google with others as well and, you, know it is a an. Industry, term. That gets used a lot and often it's well you know hybrid is not necessarily, clean it's not a pure, but it's actually reality that most people are dealing with this, survey you, see here 87%, of, enterprises, are looking at a hybrid strategy have some form, of it. 56%. Are already using multiple, types of deployments, multiple clouds, and. Only 40%, are really capturing, you know cloud first as. A, model and, even the folks that are doing cloud first often, have to go through some transition, hybrid, is a transitory phase as well so you're gonna probably run into this unless you've never built any software, who, here has no software to run. Excellent. Ok good you, have no software to run you probably have other talks you want to go to this. Is about people that have software they have to move it and and, how do you do that in a way that's that's efficient, and makes sense. So. With, that hybrid is persistent, and it's pragmatic, one. Of the one, of the ways to look at hybrid and I like, to misuse terms, so I like to tap, and other scientific, disciplines like thermodynamics. And. Entropy is a common. Term around how things when left alone eventually, devolve into a more chaotic state or maybe more, natural. Chaos and. Hybrid really in many ways is the expression of that in the IT landscape, as, your platform, evolves as you add new services, new capabilities, as you add extensions. To your architecture, you oftentimes end up in a blend of technologies, a blend of environments, a blend of services and so, this is something that naturally occurs but that's not necessarily bad news, hybrid, in fact is actually. Desired. For many reasons so. Enterprises. In general often never, single, source IT you, know that's a common common, knowledge don't put your legs in all your eggs in one basket so a hybrid strategy, is inherently, if you're gonna use multiple. Clouds multiple, environments, the. Other thing is you, get some advanced advantages, of extending, your environment, with the best of breed so hybrid is an opportunity, to say hey I've got the best pieces from this platform and I want to blend them with the best pieces from that platform and that, a functionally, is hybrid so there's some benefits, to that approach as well. And. Then. Fundamentally, embracing, hybrid which i think is you know what I'm trying to encourage.
Folks To do. Influences, how you think about IT and how you wield it so. Instead of thinking of every transition, or every, migration. As a I'm going, from A to B and B is my final state you. Recognize that hybrid will continue. To exist over time that means you adopt, approaches, that are not about n plus. 1 but what's beyond that so. If you're going from A to B what's C look like what, are the tools and techniques I'm gonna want to use going, forward even beyond this current transition how do I think about capturing, and and choosing. Tools that are gonna benefit, me over. Time and not just for this one period so thinking about that approach differently, changes. Perhaps how you think about the partners, you select the tools you choose and the approaches, you. Take. So. All in mass migrations, people. Talk about and they're fantastic it's, a great idea but, even with that over time the technology landscape, will change you will end up probably in some form of hybrid again later, after that mass migration, so, thinking about how to tackle, that in a efficient. Long-term. Way is definitely, a beneficial to you and your migrations, so. It comes in many forms as you can see here we have on the bottom. Your. Lives the bottom there you go we have kind of like a open, source e cloudy type tools here my datacenter, more traditional, bits, some. Other cloud that's not GCP and of course the Google platform so you know hybrid could be I'm blending. Two data centers with a disparate, technologies not. As exciting for this talk but that's a form of hybrid or. I'm connecting, my traditional, with a whole bunch of cloud you, know cloud first or cloud native services. Alternatively. I'm using, a common set of tools a common set of open source tools across multiple environments, here you see kubernetes. Down below gke. Up above tensorflow, down below machine learning engine, up above frightened so that's another blend, of hybrid technologies, and of course I would. In my in. My definition of hybrid point to multi, cloud as a hybrid, mental, model because you're actually blending, different approaches and different technologies, across that so, all of these really are hybrid, of a form and. So there may be things you'll run into and they're all valid, and reasonable depending, on what makes sense for you and, then, I like and. Hybrid, is is a course at every layer thinking. About hybrid it's, not a. A particular. Architecture inherently. It, generally is a blend of things and so it actually fits and there's tools that are needed across the board at every layer fundamental. To the whole thing is obviously networking, right. There's no hybrid architecture that I've ever seen that doesn't have a network in it so, kind, of makes sense so. You have to have a series of tools and you have to have a rich platform. And technology, that allows you to extend, your network as, it stands today. And integrate, other services, and then build a dynamic. Network on the other end to, support your hybrid migration so you, know in GCP we have a variety of different tools you know from dedicated. And partner interconnect for. Bridging, that gap as well as obviously VPNs. And then you know things like virtual private cloud load balancing, cloud VPN is ways to build, out that network on the clock and. That map's back to your traditional you know routers and switches that you might own today at. The next layer up you, have infrastructure, as a service. Tech. Capabilities, so traditionally, over here you've got your old you know your VMs you've got your persistent, disk you might have this, sort of looks like a very traditional legacy.
Virtualized. Infrastructure, in the enterprise. And. A series, of tools like the transfer of climate compliance, or even our new fellow strata. Tools. That, actually allow you to migrate VMs, in minutes so that's how you get those existing. Kind of lift and shift models to. Move those components, over and on you know on the on the GCP, side you've got compute, engine and cloud storage even, cloud spanner to replace maybe some traditional database, technologies. Up. Top the next layer of hybrid is thinking, about that application approach. So. In the infrastructure, layer you're maybe blending, you know some different types of services or components, at, the application, layer you're saying look I'm gonna pick a common tool chain a common, set of services like kubernetes, across the board is steel across the board and I'm gonna run that in both environments so I can actually have a single. Software. Development stack everywhere. And. That often is not everyone, goes that way from hybrid perspective, because maybe they're not in containers. Today but, this is a fundamental. Model that. You know you see with our cloud service platform announcement, you know we're espousing, this idea of a common tool chain everywhere but it does require possibly, some rewrite so, that but that is something to consider, how, do you get to that, and that's a modernization step, generally, for existing applications, and. Then, fundamentally on top what, I think a lot of folks forget about in the hybrid model is this idea of a common set of services. API. Management, you. Know visibility, in terms of monitoring across the board as you add. Services. And capabilities in multiple environments the management, infrastructure. Gets even more complicated, so thinking about a way to pick, tools and technologies, that give you visibility. Across the board like for, example stackdriver which. Can work across multiple clouds, as. Well as even your own platform or Apogee which can work across multiple environments, as well so. Those are definitely key. Elements to consider across the board. Now. Our approach to hybrid is what I'd like to talk about next but. After a drink of water I. Talked. About the network the network is the foundation, without, the network you don't have a strong hybrid story and. Ideally. To make a lot of these architectures, work you have to have a rich global. Performant. Network and that's what Google's been building for 18 plus years you, know we have 17, regions 125.
Points Of presence we, have over a hundred thousand feet of undersea cable that we've invested, in this. Is the largest private network, in the world and. We continue to build that out this is our continued ongoing commitment, this. Means, that you have access when. You use GCP with your hybrid environment, to this entire network and as your data flows between regions, between, endpoints you, have that that backbone, that we've built for our own platform and you can actually ride on top of that world-class. Infrastructure. And. We like to make it easy for you to get involved in that so we obviously have you know the Google network but. We also build multiple ways. To get in there's premium, versus standard and different network approaches, there's, also you can just do you know VPN, tunnels, if you want something quick and easy but. We also have you know dedicated interconnects, or, partner interconnects, where you can drop into a data center that you're already in and drop a simple. You know 10 gig connection across. The data center to tap right into the Google Network so. Giving you that flexibility, is key and. That is the foundation of any hybrid. Solution that you have a strong, scalable. Global. Network. The, other thing that's important and this is kind of more of a positioning, piece, around how we think about hybrid is that, fundamentally. Hybrid if it's going to be effective is it has to work in a lot of different locations it has to come to you and meet you where you're at so. Our our, principle, here and you saw this in the cloud services, platform is that, you know we're gonna make it better with open innovation, so we're gonna innovate is with, the fastest out there we're gonna build you know the latest, technologies. We're gonna give you those latest tools on the cloud but, we're also contributing, those back into the open source community so, you can use that you know the latest tensorflow. You. Know api's and capabilities, on a machine learning engine you, know on google cloud but, also you can use it in-house as well so. That's the fundamental focus, that we take and why we think hybrid with Google makes so much sense is, because we're gonna give you a common set of tools and technologies, you don't have to have you. Know developers. That understand the, Google service that you know then have to learn another service for your in-house, platform. So you can actually standardize across the board and. This makes it portable it makes it extensible it makes it available everywhere, you need to run and. Then of course there's innovation that, we continue to contribute is about not, just delivering value today but also going forward so as we continue to evolve kubernetes, we. Will do that in gke but we'll also do that you know in the kubernetes. Open, source projects as well so you're getting those innovations, across the board. These. Are some of the key things, in the last say six plus months we've done kind of hybrid space I'm not gonna read all of them this. Is being recorded and I'm seeing photos being snapped so that's awesome, you. Know a lot of things we're doing around, the infrastructure, layer or this transfer appliance Avella strata you know things like that cloud, file storage just came out partner. Interconnects, for tight, integration with, your data center to give you high throughput highly. Scalable network and then of course you know five plus or five regions we've added unto. The platform in the last six, or seven months right on, the application, layer you know support, for things like shared V PC a memory. Store service for caching, in the platform, you, know pub/sub emulators, for various, tools that are common like Kafka private.
Clusters To support your architecture, the way it fits from from a application. Perspective and then, in the management space I think you saw a lot in in the cloud services platform but, even before that we've, already been working strongly, on, integrating. Sto with Apogee, having. A Pidgey expand to multiple regions you have a way to expose that API, layer over your applications, around the globe and, then all the work and I am in terms of conditions resource management so you can not only deploy, your infrastructure, where, you need it but also manage, it in. Alignment with your own security, practices, and, principles that you have inside your enterprise today. And. Of course the cloud services platform which, you've seen who's, seen that makino, alright. You're obviously all asleep because there's on o hands somebody. Saw it you saw it you saw right that's two um four, all right we're getting there six. So. The idea here is we. Want to take what we've done in. The, in the open source projects like sto and kubernetes, and we want to make that available where. It's needed and which is not only in our the, the GCP, platform, but also in. Your data center today and. So, you know the cloud build function the cloud build and serverless capabilities, for people who are builders don't want an integrated, workflow, server. Lists you know as baked, on top of the gke platform. Jiki, on Prem which, gives you the ability to run kubernetes, both in the cloud and on-premise. A consistent, way stack. Driver and sto integration. So you actually have the ability to see API usage across. The. Your, existing, infrastructure, internally, but also in. The, cloud and then fully managed SEO is another key piece there and of, course this all integrating with the apogee API management gives you a single management plane across multiple. Environments and. So the way to look at this is I, actually, drew an architecture, diagram why, not and so you see here there's, a whole bunch of approaches, to the cloud and they're all kind of represented, now as part of this cloud services, platform architecture, so. Here you've got your your your legacy cloud nope the things that never move that's, they're very exciting over there we'll, leave them alone but they can actually be it still, connected to the cloud with a dedicated, interconnect, and then using the Apogee API platform as a way to expose, legacy. Services, with a web service interface right, a very. Common approach to help people kind of expose something that's maybe they, want a way to call into it but they don't want to be able to have to rewrite it so put an API layer over the front. And. Then staying on from in the, enterprise data center. You can use GK on Prem to modernize in place right, I want to improve my application, move to a modernized, architecture, but, I don't want to move it out of the data center maybe it's tightly, coupled to other services, in the architecture, I can leave that there but, I can add a managed, it, steel mesh over, the front over the top of it so I can actually control API, access even, though it's in my data center and. Then of course that, same approach can be applied to my remote site so for, example I might be you. Know a retail, organization, and I have you know hundreds, of stores well guess what I got to still control the services, running in the gke or the Jiki on Prem in those stores as well just like I control the services, in my enterprise data center so. You have that common modeler and then here you know in the cloud platform you might do you know lift and shift where you bring. Some things over with sort of El Estrada but also the move and improve like, I'm gonna move some features over and then improve how they run today, tying, that into the the manages to you and, what you get with you know this particular approach from a hybrid perspective, is one. You get central cluster management now you can, control. The you know kubernetes that's running across all of your environments, not just the one that's in the cloud right you have a centralized, way in which you can view how. Those clusters are performing what they're doing and upgrade, across. The across, the board you. Also have a build cloud build with a common, tool chain so as I build containers, as I build, those artifacts I want to deploy them I now have a common set of services, and api's, that I can use across the board and I, can deploy that to any of those clusters.
And. Then the other thing is fundamental, to a true hybrid models, of centralized, API management this ability to, you. Know capture a who. Can access what in my architecture, so if you go down the microservices, model the, ability to say. You know front end here can call back in there and control you. Know throughput, rates access. To those various services all through, a centralized, API. Management layer. Now. That's obviously a little forward thinking that was our announcement this morning that's kind of where I see hybrid going it to some degree and there's also a lot of other tools in space and and, we'll hear more about that but what they're what we're seeing in terms of real use cases today I thought I'd cover that just so folks can you know get excited about what might apply to their particular business today so. There's a few we see kite quite commonly I'm one isn't this idea of invoking legacy, dependencies moving pieces of my architecture, into the cloud but calling back into existing, services that I can't move today or don't want to you. Know mainframes, large-scale. Packaged. Applications, that have too many dependencies, to be picked up today things. That you know maybe run an infrastructure, that you know doesn't it's not cloud ready right but, I want to still tap into that from, the cloud that's a pretty common you know front-end. Calling, into a back-end kind of model in hybrid another. Is cross environment, execution, so I want all the benefits of CI CD I want a fantastic, dev test process I want flexibility. For my development teams I want to move quickly but, it turns out I still needed to play that code in my oil rig right. Well, you know that's gonna be a model, you're not gonna change your oil rig necessarily tomorrow so hey you, know the ability to use, the cloud as a as a cross, environment, you, know kind of execution, development, environment, is a common. Platform but deploying directly from there it. To the edge is. Common, as well and then edge computing in general I want to add a whole bunch of sensors capabilities. To my architecture, I want to gather a whole bunch of new data that I've never processed before I probably, don't have the processing. Power of the storage that I need how, can I use the cloud as a way to extend my, architecture, to capture all that information, aggregate. It feed it back into my existing application. And then use that as a sort of a blended hybrid model. Um. So some examples of that you see the GCP front end right, you, know I got my back-end services, my my existing stuff I then, have all my client traffic coming into you know global load balancers I can hit whatever region is closest to my customers, I can capture all. The front-end work that I need to do on that and then you know when I do I can use interconnects back or a VPN back into, my existing architecture. To tap into other systems like say ERP, or or back-end. You know inventory, management things like that. That's a pretty common scenario we see quite a bit might, apply to what you're doing today, the. Other one we see as I mentioned IOT, right you've got you know sure my devices are gonna continue to reach back into my existing. My. Existing, you know on from infrastructure, but then at the same time I can use cloud pub/sub IOT, core data. Flow in this whole array, of analytics. And storage services, to, capture process, and then, aggregate. That data so, then I can feed the results, back, into my. Existing architecture, for maybe you know I'll put back to my my.
Devices So I have this ability to take this maybe. Smaller. Scale footprint, and then add a whole bunch of managed, services, and data processing, capability, as a hybrid model building. Obviously a hybrid network interconnect, in between those two. Some. Other examples we're seeing is just tapping into the the global network of gzp as a way to get footprint, around the globe right as you expand your business as you expand where you your customers are coming from you, want this availability. Everywhere you. Want to be able to have the closest location we tapped into so there's global availability the 17 regions is a huge, benefit and you can go from tapping, into one region to get into that network we saw earlier to get around the globe another. Is jurisdictional. Or data sovereignty you might be tapping into a new market and there might be rules about where the data has to live has to be processed. Having. Regions there available to you right away very example but you still want you know visibility, coming back into your central, IT. Infrastructure. So, blending. Those those, data storage, and processing requirements. With tying. It back into your existing, centralized. Approach and architecture, and then the last which is quite common is this idea of extending an architecture, with cloud services, you. Know I think IOT is one example but I'll dive into another one here which is, actually. Let me go back real fast. Another. One here which is this idea of like let's say I want to add ml, or I want to add Big Data right these things are things I can tap into my architecture, and, I don't have to rebuild it I just have to figure out way to pump that data across and. Bring the results back into my platform, so. In terms of regional availability, you, know here's it you can imagine this kind of a model model, where you have a global load balancer you, know you're exposing two customers across the globe and. Then building, proxies, in the region's running. Individual, sort, of localized, services, for various. Performance. Needs but, then fundamentally, tapping that back into a and to interconnect in terms of feeding that data back into you. Know your your maybe you centralized, or lower. Lessened. Footprint, enterprise. Platform. The. Other one being the Big Data example, when we talked about you, know yeah I've got a ton of applications out over here don't fit on the screen I didn't the screams would be big enough so I was wrong about the Hat obviously, so, his hell you know zaps over here and then, that data is being pumped Pumped, through over a variety of different capabilities whether, it be a transfer appliance, you know transfer service or just, a VPN interconnect, you know whatever makes sense into. You, know an event stream ETL streaming, staging. Buckets very, streaming. And storage services, they, give you your time series analysis, your data warehouse on bigquery and then you can build a whole analytics. And reporting dashboard, over the top of that which can be exposed either in the cloud or you can even pump some of that data back for. Reporting on front so another, common, classic, hybrid model we're seeing in, the, market today um. So this is the just summary slide that shows them again there, you go. But, fundamentally we're, obviously doing a lot for the cloud services, platform and. We've, been doing a lot in terms of commitment to open. Source tools and technologies, but we also let the platform, to be open for, our partners, as well this, you know hybrid fundamentally. Means a blending, of disparate technologies and. We recognize that many of you are working with existing partners, today you. Have a wide, array of services, or, capabilities, that. You're trying to bridge, into, the. Cloud platform and so, we, think that it's as fundamental. That we actually make this platform extensible. And allow others to participate as well so you can your best choice, tools are available for you as well so. Whether it be you, know multi, cloud management capabilities, that's why we expose all of our you know api's.
So. You can actually tap into that from. A variety of tools and management consoles, hybrid. Applications, when you want to run containers, everywhere, gke. Allows. You do that you can run you. Know kubernetes on Prem you can run it with with other. Hosting. Partners as well hybrid. Services, across the board in terms of securing. And monitoring services across clouds and then, you know hybrid infrastructure so that commitment to the VMS, that can move around, you. Know V to V V to C you. Know that cloud interconnect, and data migration, their fundamentals, and obviously, we have a whole bunch of partners in the space but. What I'd love to do is turn it over to one, of our partners right, now. To. Talk more about his, story. And their solutions. Sri jiva pali who's a general manager of cloud services in botanics thanks. For joining us Matt, hey. Folks, my. Name is Sheila they were a Poli yeah. That's, you, and. My. Responsibility, is to drive forward. Hybrid. Cloud strategy at Nutanix and I'll talk about what that means in upcoming, slides and that's given you a ton of information on you know his, perspective, on hybrid and the customer, problems they see I would. Like to do the same from the mechanic, standpoint before I go any further I'd. Like to see a show fans as to how many of you actually heard of Nutanix yeah, I see okay, that's quite a few it's quite a few so, let's start there just just to recap, what what, what botanics does as a company in what what's a mission statement right. We. Deliver, infrastructure. Data center infrastructure, for customers. On-premises environment, and our. Mission is really to bring cloud. Like simplicity, and agility and cost economics to, on-premises. Data center environments, and we. Build turn, key infrastructure, stack that allows customers. To focus on their applications, instead of low level infrastructure Const constructs, we're, about nine. Years old and we've. Got 10,000, customers so far it's a fast growing company so. How. Do we how. Do we accomplish our mission what do we what, do we do to realize our goals, well. To begin with we took a deep look at data center technology choices built. And delivered piecemeal. To. Customers, and the operational, silos that they create traditional. Architectures, require to plan and procure compute, storage networking. Virtualization. All of these independently. And manage, the lifecycle of, each of these components independently that. Results, in long, lead times dedicated, teams specialized, skills, and in, general, a pretty big operational. Hazard. Right so, what have we done we went ahead and designed a new system that coalesced all those technologies, that you see on the Left compute. Storage networking, virtualization. Management, automation, orchestration. And built. A integrated. Single turnkey, appliance, form factor and the result of that is its unparalleled, simplicity. As far as operations, Center operations was concerned and planning. Is concerned and the value for customers is evident, from the, rate at which that the company has been growing with time right, having. Said that now what are we doing with, with. Google and GCP what's our joint. Vision. I. Can. Get to that as well yeah there you go our, goal is to deliver a seamless, one-click, hybrid cloud experience, in platform, for customers who right-size their applications, and workloads between, private and public clouds what, exactly do we mean by that essentially. As you, start from your on-premises, data center environments, and you try. To figure out an access path to the public cloud to realize very specific, hybrid use cases and we'll talk about more about what those use cases are we. Want to give you the best seamless, experience, possible without, having to go through the complexities, in, transformations.
In Redesigning, in refactoring. Applications, so that that's, the that's, the access path that we've conceived and we'll talk more about how, exactly are we going about realizing this, vision there. Are a set, of capabilities that we're offering to make this hybrid journey seamless, for our customers and I'll break it down into these three sort of broad areas, one. Is you know you got to start with transforming. Your own data center because that's where most of your applications are running today and what, have you done to sort of bring a degree of simplicity and agility, to that environment and that's where Nutanix as a company, is bringing value to our customers, it, were proven to provide agility, and simplicity and transform. Your existing, data center to be more cloud like that's that's been our mission all along in the success path for us anyway now. In. Addition to transforming data centers we're also allowing, customers to transform, their applications, redesigning, them to be more micro services like using technologies. Like kubernetes delivered, natively. On premises, off of our appliances that's again you know that's think, of that as Google. Services, delivered, natively out of Nutanix. Platforms, and our goal eventually is, that you, consume, a, combination. Of Nutanix and Google services in respect to of the location, they could be on-premises off premises, but, in the end you can you. Can start building your applications. In a way that, you know you progressively. Modernize, them without, having to worry about location. And transformation. Another. Key aspect which is related to what I just said is the, ability to move applications back, and forth between on premises in the cloud without having, to read again redesign, or retransform, them the. There's. An offering that's new that we call high cloud services and I'll talk more about what that means. Nutanix, stack for GCP, you can use I for DCP, integrated. With our on-premises, offering allows you to move workloads, back and forth as your, business requires. Without. Having to go through many of the you know typical, complexities, that one goes through for cloud migration right. And related, to that is essentially. Orchestration. And automation, of placement. Of these workloads and I'll talk more about how, another, product of mechanics, which is calm plays, a big part of, in. Solving that problem and, finally. Eventually. You. You want to get to a place where you refactor, the application, to optimally. Run in the cloud therefore you're taking advantage of all the native cloud services, and, not just doing a lift and shift from on-premises, to the cloud and we help you progressively. Refactor, those applications, through a through. A series, of tightly. Integrated, services, between. Google. And Nutanix. So. Those are at a high level those, are the, capabilities. That that. That we're not me designing together now. What. Use cases do they enable, you to solve. Again. As I said any use case that starts, with your, on-premises, environment, and needs to access the cloud as a seamless extension of your on-premises environment would benefit from these from, these joint offerings the, first and foremost disaster, recovery and business continuity this. Has been a. Problem. That's been it that's that's pretty hard to address traditionally. Because, it's complex in nature and.
What We've done is we've built a dr. To the cloud solution that's tightly integrated with our on-premises, offering, and with with Google and e-tanks together you, can leverage the. GCP, cloud as the target for your as. A dr target for in in the case of a failure right, a related. Use case to dr, is lift and shift as as. We talk to many customers they're looking to miniaturize, the data send a footprint and and trying to figure out a way to move. Workloads into the cloud while we can some, existing workloads, on-premises how, do we make that motion, seamless, in the lab you to still have, backward, connectivity, and access ability to what's running on Prem so, that in the end you realize that true hybrid. Situation. Where. Your. Your, actual workloads are spanning both the public and private clouds without. You having to go through again a, series, of investments, in data. Center migration, technologies, and so forth right. And. Another, use case is cloud bursting and there's, and we'll talk about a specific example where, there's a customer, who actually wanted. Additional. Capacity, in the cloud based on fluctuating. Demand from their users and. How. Do we enable that use case but allowing you to expand, into the cloud just, as much as your applications, need as though, that's that's. Sort of a seamless extension of your of your existing, platforms so these are some of the use cases and the broad theme is that any. Use case that is, gravitated. In your on-premises, environment, but needs either, intermittent. Or event-driven access, to public, cloud services will. Will. Benefit, from these joint offerings. All. Right sorry, about that there was a little bit of a technical glitch so I spoke. About the why and the what then what what why. Are we embarking, on this partnership and what use cases can be addressed let's talk a little bit about the how what, product, offerings specifically, will, help us address, the use cases that I just spoke about the. Centerpiece, of it all is actually to enable a seamless hybrid experience, the first thing to solve for is to keep the same operational, paradigm the same interfaces, the, same tooling, the same workflows, to, manage your applications independent. Of location, and there's, a product offering from Nutanix called comm which stands for cloud. Application, lifecycle management, and it, allows you to manage the lifecycle of your applications, across multiple clouds, and by lifecycle I mean application.
Deployment Upgrades. Expansion. Visibility. Security, all those elements that you would take care of as you design and run your applications, Khan gives you a consistent. Interface to, perform all those operations, across, different clouds and. The way you would do that is you would create a comm blueprint, think. Of that as you. Know a set of instructions, that describes the constitution, of your application, and its, infrastructure, requirements and dependencies, and once, you've designed and built the blueprint, you can repeatedly. Deploy, it across different clouds while maintaining the same operational, consistency, in. The way you manage your applications it. Could be you know the target cloud could be private cloud on Nutanix, or public, cloud like DCP, and we. Come, allows you to manage not just VM based applications, but also containerized, workloads as well, comm. Blueprints also allows you to provision. And manage kubernetes, with to a control. Plane through a through, one-click, experience. Now. That's a little bit about you know homogenizing. Consumption, into operational interfaces, and that's what comm enables. You to do but. A true hybrid stack goes beyond just consumption. And where the need is to ensure that all aspects. Of your workload stay consistent, on. Both sides including things, like performance scale, availability. Security, data protection things. Of that nature. This calls for a, different level of building. Hybrid, and by, that I mean it's, the same platform, that needs to extend between your on-premises, environment, and the, target, public cloud and that's, where our, next. Offering which we, call zai cloud services, accomplishes. That's exactly what it does think. Of it as a native. Extension. Of. Of. Our net of our in a price cloud platform, into, the public cloud into, various public clouds including, GCP and. It's. Our technology stack it's the same software, technology, stack that Nutanix has built that we deliver on-premises now it's delivered, as a service in or in various public clouds including, data, centers that we run in to operate but also within the GCP regions with. Zhai not, only do you maintain operational. Consistency, as you expand into public cloud but you retain all the, applications. SLA s and. The security model and performance, characteristics, and everything. Else that you would have carefully designed, as you built and run your application, right it's, because and you could do that only because it's the same platform on both sides and the, first service that we are offering with Zhai is disaster. Recovery as a service and. We plan. To expand, the, set, of use cases to encompass other hybrid, scenarios, as well and. Specifically. With Google we are offering Zhai services. Within a DCP, region co-located. With, low latency access to the rest of the GCP services so, if you're a customer. Who is looking to transition to the public cloud for, either your existing, workloads or to build new workloads, you. Can leverage a combination, of mechanic, services, and GCP, services through this joint offering. Now. In addition to proximity, there's a few more things we're doing on the on the interface, consumption, side to make the whole experience seamless for. Our joint customers I'll, talk through some, of those. Yeah. As, I get the right slide a. Lot. Of this is related platform, integrations, to enable and ensure a seamless experience so firstly. It's, going to be an integrated portal there's a common portal from which customers, can consume, a. A combination. Of mechanic services and a select set of GCP services, that we will make. Accessible. Through this portal and those. Select, set would be the, set of services that would make sense to consume, from within Sai things like big data bigquery. That. Can actually act on data, that's the, your data that's residing in within, the Sai cloud services, anyway. Another. Common problem that customers experiences, management, of credentials, and through a single, sign-on experience we. Allow you to maintain a common identity, to. Access a combination, of against I and Nutanix services so that you don't really have to manage a different. Set of credentials for each of your employees or operators, who are. Managing. And administering these these, resources. We. Consolidate, bills and you. Get a single bill from loot annex that constant, that captures, both charges. Related to DCP and charges, related to the. New tan excise services, so you're not dealing with two payment systems and, to. Two. Ways of. Tracking. Tracking, your bills and things of that nature. And. Finally, this is important and map touched upon this before also which is network, connectivity and establishing. Network connectivity is really usually. Very cumbersome and painful we. Unify that in the sense that you, could use a single, network.
Connection To access a combination, of both sigh services, and DCP. Services and that connection, could be a VPN, connection over the Internet or it, could be a you, know DCP, interconnect, as a way to reach a combination, of you, know DCP services in Sai services and if, you see all of these sort. Of capabilities. Together the, goal is to enable. A more seamless simplified. Experience, to, access. What, we call a, set, of hybrid services. So. I spoke, a lot about you. Know essentially. Various. Capabilities, and mission. Statements, and use cases and so forth let's bring all of it together. Using, a actual, customer, example, here's. A here's. A real scenario, where a where, there's an existing joint customer, of mechanics, and ng. CP and, they were interested, in essentially. Procuring, exists, compute capacity on, the on demand as in when their. Application. Needs. So during peak seasons it's an application that serves both internal, and external users, for that for that organization, and during, peak season they need, capacity. For that application to expand on demand it's a pretty standard web, application, that that consists, of Apache web servers and Oracle. Databases, and, JBoss, application servers. And, primarily. They wanted the web server capacity to expand on demand without human intervention and, without, any kind of lead times and without over procuring capacity. Now. It's. A pretty standard, cloudbursting, use case and I'm sure many of you've heard of this use case before. But it's pretty hard to accomplish without, the, right degree of automation, and. Without, having to read transform, or change the nature in which the your applications, run especially. If you are looking to access the public cloud as a way to procure, that additional capacity. So. We've. Implemented, or in the process of implementing a solution for. This customer, that includes, a lot of the components that we spoke about at. The, com acts as the overarching orchestration. Layer and it's, constantly, monitoring. Performance, util and utilization, of the application, and. It's. It's. Built to, orchestrate. New web VMs in DCP, and this is actually either within Zhai in DCP, and. Therefore. If you need more capacity on, the web without, any kind of operator, intervention the, system itself detects, that okay it's reached a threshold at, which point it needs to spin up these instances, and it launches them it adds them to the web farm it does other forms. Of automation things, like assigning IP addresses, building. Connectivity, back to the on-premises. Environment, with the actual database runs and modifying.
DNS Records and things of that nature all, the automation, tasks that you would otherwise manually, do right, and we. Have we leverage as I was saying that common network connectivity between. GCP. Resources, and Zhai resources, as a way to establish. Connectivity. Between. On-premises. And the cloud environment and there, you have it essentially. This. Is how we've solved, this, customers, use case and if you see it's, the fact that zai is running, inside GCP, that this this. Automation, has become a minimal in the sense that the customer's workloads, need not have, been transformed. In any way, there has been no redesign, of their applications, the way they've. Constructed. Their virtual machines and the way they've designed and characterized, performance, for their applications, on their. On premises platform, all of those characteristics, are retained as these, instances, burst into the into. The public cloud. It's. A pretty simple example but it illustrates essentially. The power of the giant solution because it addresses things like automation. And the and seamless, expansion, without, the need, for refactoring. Your, applications, right and the benefits to customers is really the, simplicity that brings to the table and the, economic, advantage of adding. Capacity only, when you need it right. Guaranteed. Availability, and these are you know this is a pretty business critical application. And therefore, guaranteeing. Availability, and performance SSL is for this application while, you're taking, advantage of the economic elasticity. Is critical, right and it, also illustrates, the way in which workloads, can be moved I showed you a cloud bursting use case you can clearly, see how this can be extended over to a mobility, use case where instead of bursting, into the cloud you could actually move that entire application. Over. Into the cloud using exactly. The same constructs. And the same tools that we built to address this, use case anyway. So. Hopefully. That gives you a sense for the why and the what and the how along. With the actual real example. That. Illustrates sort of what we're bringing together between. Nutanix, and Google. So. I will be around in the night annex booth. In the conference so if you have any additional questions and, myself and my colleagues would be happy to answer them thank. You for your time and have. A great afternoon Thanks. Appreciate. I, don't, think we have any more slides but you know I think the the point that we're trying to make here if I can wrap it up is. Really, the, our, our goal at, Google is really to provide you a hybrid, solution that, meets you where you're at like. Depending on what you need for today's, approach it fits your business needs it can you know as you've seen it can move existing applications, you can modernize and improve you. Can. Networks, you. Can provide all the tools and capabilities to meet, you where at and and help you meet your business goals and what. Is the most efficient, fashion for your, business, going. Forward as. You go into your cloud transition, with. That we actually have a twist ending to. This movie which is that we actually a time for questions I. Didn't, think we're gonna be able to have that but you know it turns out either I talk too fast or, I missed the slide so. There are microphones on, that side and this side or if you want to scream it at me because you're in the middle and I can repeat it that's an option too but we have five, little over five minutes if you have no questions and want to run away I get it.
Thank, You very much for your time and we. Look forward to seeing. You at the conference oh and, don't forget if you are running away tomorrow. And the only reason I'm plugging this is because I'm proctoring, cloud. Hero it's. An interactive devops, game if you want to like literally build an, app real time for, a challenge to, see who gets the highest score there, are prizes and there. Are capes actually, oddly enough you should look for us. You.
2018-08-17 16:14
