Ready Set the Future is Hybrid

Show video

[Music] oh hello and welcome to ready set the future is hybrid my name is phoebe go and i am hosting a very exciting panel today with some recognized thought leaders and technology advocates from around the world but let's introduce the topic first we are going to be talking about hybrid cloud this term has been a while around for a while but i've seen the meaning and the purpose of hybrid cloud evolving over time initially it was more about transforming it by connecting on-premises data centers and servers and storage to public cloud resources maybe to burst or to leverage elasticity of the cloud as an emergency measure the dream to integrate these resources was difficult at the time because you needed middleware complex integrations networking and to build a control plane to manage monitor and optimize all of those resources today though hybrid cloud has come a long way and the architecture is focused less on these physical elements and more about the portability of applications and workloads and even you know user experiences across all of those all the different clouds that are available and automating a lot of that those resources and the deployment of them so that we can focus more on bringing good business value to our customers so speaking of portability hybrid also means the ability to mobilize applications to the right place at the right time maybe there's reasons like lowering cost or to speed up innovation or maybe as we've seen in the recent shift to virtual and working from home to move workloads to more accessible locations anyway in this continu in this community roundtable we will uncover what is fueling and hindering hybrid cloud adoption based on the netapp report the future is hybrid and the survey results and feedback from our own customers on their hybrid cloud journeys with their adoption their use cases workloads and business benefits so let's get right into the into the introductions of our panel folks as i introduce you since we are talking about hybrid have a think about you know what kind of hybrid would you be if you could combine any two technologies so let's start with our special guest nigel poulton nigel is a techaholic with a passion for demystifying exciting technologies he's written several best-selling books and video training courses and helped well over one million people take their first steps with docker and kubernetes prior to working with containers nigel did everything from changing backup tapes resetting passwords designing and managing hybrid clouds and more so welcome nigel hey cheers phoebe always a pleasure and honestly what an epic topic that we've got to talk about tonight i'm looking forward to it oh thank you and our featured panelist today is darnell fatigati a senior product marketer at netapp focusing on hybrid cloud darnell brings a wealth of experience in product marketing at focusing deeply on solving enterprise user needs challenges and opportunities hey daniel hey phoebe how are you any thought on your your hybrid technical specialty oh wow um i think i would have to be um a mixture of youtube music and like some sort of you know flying device like maybe a jet pack or something i don't know it sounds like fun i love that that sounds like fun up next is matthew underhill he is the head of infrastructure at alfred h knight out of liverpool matthew's broad experiences in the industry focus on supporting businesses to achieve outcomes by delivering robust and reliable infrastructure matthew is also an enthusiastic member of the netapp a team hey matthew hey phoebe thanks for having me any thoughts on your technical hybrid well when you said technology i was thinking there's a lot of things that could be technology so i think i'd like to be a fridge mixed with a laser because everyone loves lasers and then i'd be cool you'd be a cool laser next i'd like to introduce karen lopez karen is a senior project manager and architect at info advisors she has 20 over 20 years now right of experience in project and data management on large programs karen specializes in the practical application of data management principles and she is a microsoft mvp specializing in data modeling and database design and welcome karen any of those your technical hybrid yeah sort of so i was thinking about this and i think my combination would be vintage sports cars and something like peloton where i am combining old things like my body and all the data i can measure about them oh i like that that's excellent and you could maybe have in the sports cars your peloton like you can do it you can yes exactly exactly and last but not least michael hay is the vice president product management of private cloud and business continuity at teradata as a strategic and recognized technology and business leader michael's efforts have been instrumental in driving innovations to produce significant and sustained business outcomes hey michael welcome to the panel how about you you've had a little bit of time to think about i have and so i've got two right yeah because i've had a little extra time first is sort of a little hobby i've been developing which is somehow combining like old vintage computer chips and art so i'm kind of doing some pour overs and i've got some barren walls i'm trying to decorate so that's kind of interesting kind of tactile and hands-on another something i thought of for a long time which is how to combine like agribusiness and i.t in some sort of weird quasi-economics way so a little bit of a brainy thing a little bit of a tactile thing that's how i'd hybridize so really great to be here by the way super happy to uh entertain the discussion yeah yeah it's it's gonna be a really fun one all right so let's bring everybody back up onto the screen and let's just talk initially about what is hybrid cloud so before we get into the study results and some of these topics let's kind of think of an a definition so darnell um you know hybrids a lot of things for a lot of people as we found out but how do you define hybrid cloud today you know i think actually your description you got um you nailed it and you know i i look at what nist defines hybrid cloud to be so the national institute of standards and technology and basically they define it as two or more distinct cloud infrastructure so it could be a private public that remain unique but they're bound together by standardized or proprietary technology that enables as you said data and application portability so i you know i think um you know as you mentioned a lot of different people have different definitions you ask 10 people you'll get 10 different answers but i think the real key thing here is that they're bound together and uh it's enabling accessibility of that data okay yeah that's interesting you say they're bound together do you think that's changed from an um you know how maybe we thought about hybrid cloud when we were first kind of investigating this architecture yeah i mean i i think that people still define hybrid cloud as two disparate or two or more disparate groups of infrastructure i don't even think that you know there there is a standard definition that people have like fully embraced and adopted yet because i think you know and based on some of the questions in our survey i think initially we wanted to make sure that people understood what we were uh talking about and how we're defining it but i think initially they defined hybrid cloud as being separate and not necessarily integrated so that that's an important distinction yeah so i know that darnell said like it can mean anything to anybody and maybe everything to everybody and she talked about portability of workload she talked about um standardized components as well there's if i include you phoebe there's six of us on the panel i'm just gonna throw this out there right among the six of us right we can all come to a consensus that any kind of hybrid cloud to to bring that portability and to sort of knit the desperate infrastructures together it's got to be kubernetes right yeah i know it's not like it's not an animal standard come on because you know you're kind of heading down that path yeah exactly like for instance is it really hybrid cloud the way we intend it if all i'm doing is sending my third copy of backups to the cloud not my first copy of backups not my secondary but my third is that hybrid cloud i'm not sharing workloads so i would say current i totally get where you're coming from there and it i do think that unfortunately that's hybrid cloud today but it probably won't be going forward data's obviously been one of the more difficult um challenges for any kind of workload and virtualization and hybrid of anything right so i do feel it's really early days there um i kind of wish that wasn't the definition but for me i think you're dead right i think it's not great but that is part of the definition of hybrid cloud today yeah and i do think you're you're on to something there nigel because how to get started right there's operational stuff that goes along with the portability whether it's computer data portability and the way you exercise that is just starting from somewhere if it has to be sort of that third copy and eventually you go from just a static dark copy of data to something that's active we certainly hear some of our customers talk about that you know let's get the data there and then let's activate it and it's it's kind of a two-step process and i i think um you know darnell and phoebe just to react to the whole i've got vulcanized clouds right so balkans was the notion of hybrid that has to be shifted towards active right it needs to be whether it's a simple backup whether it's active data whether it's something more a new workload all of those become germane to kind of shift the definition for sure yeah we're definitely seeing some really cool conversations where we used to have you know primary data center and maybe a secondary or backup or a dr data center and now that we're starting to use cloud in that conversation we're getting customers saying i want to actually use my backup data i want to activate it i want to start doing development or testing on it or i want to do some you know analytics on it so then that kind of becomes a primary in a sense and maybe it needs its own backup somewhere else maybe it goes back on premises but yeah that's a really interesting conversation i really like that one with with customers i i'm going to just put it out there and say for me maybe it's because i'm a little bit younger than everyone getting started a little bit later i think it's an idea or it should be an idea it's not one technology or the other it's a combination of technologies and once you have that idea moving around see where see where it is so you know kubernetes could be one of the methods that you use to create your hybrid cloud but i think generally if you start off with an idea of hybrid cloud i want to have things in more than one location [Music] that will help you to really achieve hybrid cloud rather than starting off well maybe you get you know something from the sea level we must have a hybrid cloud so we squirt off our backup and then that's what we're calling our hybrid cloud kind of thing yeah i still think some of that's there matthew i mean it's it's uh by fiat by order of the executive we must have the cloud right that's yes that's a key part of the definition but yeah i think there's you know what's the purpose um i know the u.s patent trademark office has a free data pool and you know if you wanted to do some competitive analysis and mix that up with data that you might have in your four walls um that seems like a righteous use case you get back to purpose and value um and if you know if your definition of hybrid doesn't include purpose and value i don't know what the point is yeah that's very true well darnell in in that survey um yeah what were some of those common workloads and use cases that customers were you know respondents were kind of talking about using a hybrid cloud form well and it's interesting to hear you guys talk about this because it sort of reflects what we found in our survey is that data protection right so that it's readily understood it's the most widely uh cited use case for hybrid cloud today because you know we're talking about backing up um you know secondary copies creating dr uh locations even archiving um and and tiering probably second to that um but i think that you know customers can see the value of that immediately right like it's much more tangible um the tools are readily available um it's it's easier for them and one of the biggest challenges that they have in executing their hybrid cloud strategies is around skill set right and and the the the choice of tools out there and the disparate tools out there so you know starting off with something like data protection uh makes it a little bit more it brings it down a level it's it's you know easier um it's easier for them to get into and it's easier that for them to see real benefits right away so that's one for sure and that makes me think right karen's worked a lot with data we talked we've talked about data a lot and so i mean like you said backup is kind of that we all get it we know what it's there for we know what you use it for but karen in your kind of experience when you're talking about hybrid cloud and data what's your we're having a really fun session with lots of cool background noise that just makes it fun we're real we're live what are your thoughts karen well i think the main thing is is that what forms people's decisions today on hybrid cloud are the costing models and because it's cloud just be the technology's agile we're more agile but how things are billed for how they're charged how much commute co compute costs all of those things are impacting our technology and architecture decisions more than ever before because we're really going from a very capital expense driven architectural process to an operational one and i think if we had this webinar two years ago or two years from now we'd come to the same sort of conclusion that the right answer is going to change over time also you know if the cloud providers were ever to start charging for ingress that would also change these types of decisions so it's kind of funny how you know follow the money also works with follow the architecture yeah actually karen i want to just um touch on something that you said because we saw by far in our survey that in our survey that the number one driver for hybrid cloud adoption is flexibility so i mean that's it right i mean your your cost requirements change you know the your your risk requirements change your reliability and performance requirements change and i think that that's you know why customers are saying right now that they you know want to create these platforms because they want that flexibility well as as people who have built platforms michael or matthew what are your thoughts on on that concept um yeah i think the first thing is don't fall prey to fashion uh we in the tech industry we don't have runways you know we don't have like fashion week but we do have resumes and there's a tendency to kind of chase around um a buzzword and then you know the the trough of disillusionment all the stuff that some of the analysts teams talk about is what the journey you actually have to go on and i think to karen's point you know 24 months from now when the the doldrums of maturation have sort of sunk in you know that that's when people will will be talking about not hybrid cloud they'll be talking about the value that they have drawn from flexibility the value they have drawn from the use cases right they were able to say mix up since i talked about agriculture data and patent data and figure out something that you know insulated them say from you know the the inclement economic circumstances of the day right so that's what they'll be talking about they're not going to be talking about did i get my third backup or my archive in the cloud they'll be talking about the value that has been derived and that's where we need to get to and so i think the thing is to understand the fashion but don't fall prey to the fashion that's really important phoebe i know you're going to come to matthew in just a second i just want to pick up on that though michael i was gonna think yeah oh no you're gonna talk about kubernetes aren't you sure um i do feel like you kind of saved yourself a little bit at the end there michael because i was i was concerned i was concerned with the idea of not following fashion uh i i've like i fully appreciate where you're coming from but when the writing is all over the wall for something and and we're constantly making bets in technology you know almost every day which cloud do we go with which platforms which orchestration systems do we go with um with an eye to the future that like we want to invest um time and money and skills and talent in things that will be of lasting value for the organization and lasting value for customers um i use an analogy a lot and i don't it's a little bit cheesy but if you were an organization that decided virtualization like in in the guise of vmware or whatever wasn't for you as an individual or as an organization you know made that decision 10 15 years ago or whatever you would likely be out of business or out of a job as an individual now um because these waves have to be ridden and when something as big as vmware or as big as and i won't say it right but as big as some of the waves that are coming along at the moment are if you don't ride them as an individual or as an organization i feel like there's a hefty price that you will pay further on down the line so i i don't know if it's if it really is a thing but we used to say when i was growing up fashion fades but style remains yeah and i don't know if yes we shouldn't watch like the the fads that come and go but if we see something that we feel is a bit of a juggernaut or is you know something that is almost a tidal force that we've got to roll with it um otherwise there are serious risks for the organization and business and and you know for us as individuals if we want to get a job in the future i i want to come back to you to the technology part but i want to actually ask matthew as somebody who's building platforms right i thought that like the comments that you guys weren't making was around well don't follow a fad but part of the thing that's important with following your fad is what skills do i pick up or what skills do i train my team on so matthew when you're kind of looking at this and and kind of these building these new architectures hybrid cloud architectures what kinds of skills are you looking at enabling your people so that we're able to ride these waves but not fall prey to you know ugg boots and bad hair well actually i mean i was going to just respond to nigel but in terms of the skills and things i think you know as tech people and as i'm looking for recruiter recruits and people to join my team you have to be looking for people who are interested in the latest technology um you know we're all going out there getting our our various certifications or or you know just being generally interested in things as they you know grow and as they progress but i i would say in terms of what we're actually doing as we're building you know we are well alfred h night we're 140 years old um we've been family owned and family run and that's kind of kept us very stable you know very grounded and we're not looking at technology in our production that's you know not it has to be well proven in the field otherwise you know we're unsure is it going to break is you know someone going to release you know a patch and then there we go um and so yeah you know we are looking at the new technology but we're also building our things on technology that has been proven in the past um because otherwise otherwise you know um well it's my head on the chopping block i guess well i mean i just i think that's a really important point and uh you know perhaps if you're a leading edge cloud company being on that bleeding edge is important but you know if the intensity intensiveness of your business uh has to do with not intellect information technology uh you know you don't need to be on the latest bit you need to be on the stablest bit because your business depends upon getting to value and insights i mean i have a hobby of kind of working on some data projects on the side and frankly you know two-year-old docker technologies just fine because i'm interested in the business problem so i support that 100 percent so most of my experiences at enterprise organizations and the most amazing part about it is when i'm hiring of course i want someone with the skill set that we need but i want someone that can actually do cost-benefit risk analysis about change and i work with i mean i'm kind of this way too as a gadget person that you know i'm excited about new technologies but that doesn't mean it's going to be a good fit for the problems i'm trying to solve in enabling a business to sell the widgets that they're selling or developing the widgets that they're developing so it might not be today karen but but i feel with with anything that's like a sea change like virtualization and cloud and things like that that increasingly it will be the the resolution to your problem tomorrow like i know i don't want to sound like i'm banging the drum about kubernetes because you know i've spent probably the last four or five years saying it is not for everybody it really isn't um but i do find increasingly now i'm like oh it's becoming more and more right for more and more people and there's a trend going there now i'm not saying be on the bleeding edge like i i appreciate what you say karen and michael as well that like you want to be unstable you've got to deliver and value and stability today but you must have at least you know as an individual or an as an organization people or half an eye on on what the future is you know what's coming down the line because we said at the beginning didn't we that you know today what hybrid cloud looks like today will not be what it looks like tomorrow and and we don't want to be bitten by that so yeah let's talk about that a little bit that this is i mean this is a hybrid cloud discussion and so adopting these things kind of incrementally seems like it's got more benefits than saying wholesale hey we're going to a kubernetes micro services driven architecture because that's the hybrid cloud right so i mean you know anybody have kind of thoughts on um how you might adopt pieces of this where does it make sense to adopt a hybrid cloud architecture maybe with kubernetes or maybe without kubernetes in terms of that sort of thing i always think back to what matt watts sort of wrote about and it was you know cloud first question mark i guess um and that's how i approach things right like so it could be cloud first question mark or you know kubernetes question mark or you know all of all of the above and i mean you know i'm definitely seeing use cases coming down the line where you know kubernetes or containerization is going to be something that we're going to have to do but you know i think it's always about making sure that it is right for the technology so that you can realize you know the value of it rather than just saying oh you know what i'm gonna stick um whatever it is sequel or something into into that container oh that's that's broken kind of thing um i think it's you you make a good point like i think it's like i want to just tweak one of your words and it's like picking the right technology for the right workload as opposed to the the underlying infrastructure and we all deciding that this is the right kind of way to build something it's well actually what are we trying to do is it right yeah absolutely absolutely karen or michael do you have any experiences um you know that you could share with us around doing that i certainly have um so we have customers with very large systems like i've been over the past two weeks talking with a couple customers about three petabyte disaster recovery problems right these are these are enviable large-scale avant-garde data protection issues and they're thinking of hybrid cloud but they're not just thinking about it for data protection they're asking themselves uh to matthew's point like okay i have a new use case it's running the cloud what can i do can i reduce my if i'm a bank reduce my cost to comply are there new workloads i can sort of work on to discover if i'm on oil and gas and so some of that speaks to the virtues of uh the mix of public and private cloud technologies because you can spin things up but then you need that interlock function and uh the company i work for teradata has really fabulous technology to kind of build a a data fabric between uh things that are on-premises and things that are sort of in a public cloud thus manifesting a hybrid cloud and you can explore new workloads you can explore new data sets and look to solve new business problems so i see certainly um that intermix becoming increasingly relevant and it's the extension from you know hey let me create that secondary copy let me activate that secondary copy and then let me explain these new use cases that you maybe can activate different lines of business workloads and new customers inside of if i'm a user like my customers internally right yeah and actually you know michael that's pretty consistent too with what we've seen i mean in our survey and just you know from like some of our case studies and and win reports is that customers you know as i mentioned earlier data protection was pretty much the number one use case but we're seeing things like bursting much more for additional processing you know like big banks that you know need to run reports and but they only need to do it you know once per month so you know they're they're cashing data in the cloud and you know using um those additional cloud resources you know we're also seeing more um you know some side-by-side production so like you know they're essentially migrating workloads so they're taking it you know from on-prem and then they're putting it into the cloud for you know whatever reason to put it closer to the users you know to put it closer to applications that they may be you know taking advantage of cloud provider uh services so i think what we're definitely seeing is is the expansion outside of just you know the data protection uh workload for hybrid cloud but you know to nigel nigel would probably be happy to hear this we actually surprisingly when we asked about workloads in our survey we did see that um about 30 of our customers said they're using hybrid clouds uh with containerized applications which uh you know i i we didn't ask them specifically are those kubernetes environments but i mean i think we could pretty much guess that they are i think do you know on that though darnell is that i know why we've been talking about how do we um and how do organizations choose the right technologies for the right jobs and things at the right time and what have you i do feel like just about every organization though really wants to be able to standardize on something so that um it's not a case of you know we're running five different stacks for five different use cases what have you ideally an organization wants to say look we deploy this as our standard infrastructure of course there will be edge cases but generally speaking you don't want edge cases i'm not look at years ago when i was hiring for a linux role i remember um interviewing certain people from a particular organization in london a large financial organization that kind of baked their own linux kernel and all this stuff around it and it was like you know what who wants to actually be in that business when you know these days at the time linux was really stable and everybody else was just using vanilla linux like why am i going through the hard work of baking stuff myself organizations want like you don't want to invest time and people and money in learning all these different technologies let's just standardize um especially the lower level components the software and hardware infrastructurey bits um and i know it at first like you sometimes you're sacrificing performance or you're sacrificing some features in the early days but as these technologies cloud and virtualization and containers and things as they mature and evolve i mean hyper-converged infrastructures that well it was all it was always originally a trade-off of you don't quite get the performance or this and that but but now you kind of do um so you know i just everybody should ultimately i could be getting shot down here right but should be aiming for a simplified standardized stack and not and not a bunch of different silos we know that right yeah and i think that's the bridge nigel for what uh matthew was talking about right that give you that sense of normalization because investing in variance for variant's sake doesn't actually help at all right it just means you're you're kind of churning churning and not progressing progressing right so that gap of value is hindered by variation and i think that's a great call out actually right that standardization is critical yeah yeah and it's a big call out to people managing their careers because i've seen it even in the database world that the people who specialized in doing the very very lowest layer of tuning queries and applications and workloads who specialized in file level block sizes and that stuff there's less and less demand for those skills and so like there's one friend of mine who i know is very good at that and i'm like great in five years you'll be like the cobalt programmers making two quarter million dollars a year but it's not a good career plan to hope that you end up being so obsolete in your skills that now you're in high demand because everyone else has retired well just you have to go work for the cloud providers because those are the people that seem to be bottling on the talent up right yeah well this is an interesting discussion about maturity right because i think containers kubernetes even hybrid cloud is itself as a concept has kind of gone through that iterative we're going to change everything the definitions change weekly to the point where it's kind of becoming something we can have a conversation about and it's still relevant in a month or two or three right and and i think that's where we want to be with any of our architecture our technology decisions um but danielle i want to ask you know in your research how has maturity of hybrid cloud um been perceived by you know our customers or your respondents really survey respondents um you know this actually was something that took us by surprise is that the majority of respondents said either they were traditional in the sense that they haven't yet created what we were talking about as our definition of hybrid cloud integrated hybrid cloud right so they had sort of silos um pools of resources and and then the next group um which was pretty much the other half said we're maturing so we're just starting to do it which is consistent with everything that we've said right like you know they're going to start with some you know easier use cases and then get the benefits of that sort of understand the tools that you're going to use understand the processes that you're going to use understand where you could standardize things and then expand that out to other use cases over time i want to dive into that a little bit and and this is to the to anybody really is what are those steps like how do you get from group a to group b i want to be in group b because it sounds like the more progressive place to be career-wise but also you know in terms of being able to move faster so what are your thoughts on um how do you do that and it's not just you know start using kubernetes right nigel yeah wait so first you have to make 100 mistakes and then you're good that's right that's right what are those mistakes what do you think they'll go through um if they do it maybe how do you fail at hybrid cloud is it possible most often shift with no re-architecting and no um sort of dealing with security differences and how things are tuned like there are some really small workload applications that i think with low low demand and low availability needs that people use as proofs of concept and those things and that's all good but it's not going to teach you the lessons that you need to learn from a massively large workload with big data streams and all this other stuff so i'm really joking saying making the mistakes it's just if you don't like um you know the quote that goes around that cloud isn't another location it's a new way of doing computing and people who don't realize that often are going to make more mistakes yeah i want to jump off of that i think what um what karen just talked about is essential if you don't start in effect with a clean sheet of paper then you're not going to pick up the clean parts to start from that are new so if you're approaching it from the brownfield you will very much worry about lift and shift uh operationalization of things you've already got all of the stuff that you've maybe done a good job as an organization forever but if you are starting with new workloads if you are starting work there's new application potential um you know the the freedom of experimenting with the application right so it's it's a it's a high risk high return environment also carries with it the freedom of experimenting with the infrastructure and the linkage between uh private and public to make a hybrid type scenario but again the brown field that's where it gets trickier uh we certainly i i would say our customers are aggressively moving in the cloud and they're just being smart about how they're approaching it with a mix of brown and green field type applications so yeah i would say i mean sorry i was just karen took all the words right out of my mouth but uh so i definitely agree with that um you know i guess what i do is uh we architect for the cloud you know um i don't lift and shift uh i've tried it doesn't work i've had those hundred mistakes uh and then you know you bring it back and then everything's great uh so yeah architect for the cloud plan your workloads to use cloud technologies you know don't say well just yeah i'll stick a dc up there it'd be great and you know we'll be printing in no time so something that's really good actually but it's something that i think a lot of organized organizations have done like a tried and tested pattern is to take a group of superstars within your organization like a swat team for one of a better expression you know people that are that are keen to live on the edge and pick up and learn something new and make the mistakes that karen talked about but they tend to start with a net new application like we talked about brownfield i mean don't go there for your first few applications for sure and it but anything is you know sometimes something net new that's maybe internal first architect that around the cloud with containers and and maybe kubernetes and things like that um and and have that team learn on the job and document it then go something net new that is production because that's very very different production customer facing and have this team sort of skill up and start to document how they do things and then and start inserting them within other teams around the organization as you sort of kind of spread the knowledge within the organization and it's a journey and it's a journey that probably is never-ending like i mean a lot of these organizations that i've spoken to right some of them st so they'll have like containers shoulder-to-shoulder with virtual machines shoulder-to-shoulder with physical service showed at the shoulder with you know i series and mainframe these things never go away and we're constantly sort of iterating on this but but the idea is you know you get some clever people that are really up for the challenge you start with net new stuff give them a few projects they learn they make a bunch of mistakes yeah karen and then then they start disseminating that you know further and wider within the organization and i've seen tons of people you know they're not super far down that journey but but they've they did that with docker in the early days they're doing it with kubernetes they did it with cloud you know when cloud was like what do we even mean when we talk about cloud i like that because you know it really lends itself to that concept of picking a standard model like if you get too many really smart people doing their own thing and working out how to do hybrid cloud you'll end up with five six ten hybrid cloud models which don't necessarily align and then you have to kind of bring them all together at some point hopefully um or work out how you're going to integrate those to daniel's point so um i i want to kind of shift onto a topic though that is just more of a i guess where do you see the biggest advantages of adopting a hybrid cloud um i just want to get like i guess your perspective you know we've talked about some of those challenges all the mistakes you're gonna make oh my gosh so why should we do this so maybe karen let's start with you the hundred mistakes leads to well so again i come back to the costing of it all right so hybrid having hybrid solutions allows you to optimize for all the things so for the physics of data i mean some people one of the things that gets people in the cloud is they realize they can't tune their way around physics so the cloud offers you know you can take data closer to people and workloads closer to people if you have global offices and help them with performance but you're still not going to beat the speed of light when you do that the other is that trying to um trying to optimize one type of cloud architecture that is standardized across your company of course we still need to tailor and customize for the workloads for the risk for the data sovereignty like where can the data be and data gravity does it make sense to move this data around or does it make more sense to bring the compute to the data all of those types of things are much easier to do in the cloud but there are still going to be cost benefit and risk analyses that say no staying on prem with our traditional architectures makes sense right now today it might not make sense in six months or a year from now but today this is our optimized solution michael i i know we had a discussion a while ago about how some of these cloud architectures didn't just want the right decision you know what's your kind of how do you see the benefit of hybrid cloud you know now having been through that um again it's it's about uh there's different paces of friction in your your it infrastructure and uh the pace of friction is lower right so the the amount of things and just what karen talked about like you know data gravity uh the fact that there's uh more consistent ubiquitous pipes there's cdn services there's a lot more that you don't have to build and you can take advantage of and so the friction to do all of those things is like gone and therefore it lets you get to that clean sheet of paper faster and so where you can find the alignment of your needs so hence value extension with lower friction and by the way high friction is fine sometimes like compliance workloads and privacy and other things with high friction that's good um but you know innovating and you know trying to figure out oh my god you know where are my customers in this journey of this new device that they've just built and put out there i just saw that you know was it comcast has put out a new um uh a set-top box like a little thing well they probably need some data about that they're probably going to be using the cloud they probably shouldn't build their own data centers right um which is the antonym of sort of matthew's point right they probably should put the data center in the cloud in this case so i think again it's about the pace of friction that's awesome i like i like that comment all right well i think we actually have to wrap so i love this conversation and i want to keep it going on twitter so make sure you you yeah we'll start a thread on twitter and anybody can join in and if you'd like to find out more of this great content and conversation around cloud hybrid cloud private and public clouds and also how to build cost-effective cost optimized data centers check out the rest of the netapp tv content we'll be having more of these round tables we'll be having breakout sessions and we will also be um posting some really cool tech talks so i and demos as well which i like personally so thank you all for joining us thank you to the panel let's bring everybody back so i can say thank you to everybody um i really appreciated your time today and this was just such a fun discussion i am phoebego this is ready set and i will see you next time [Music] you

2021-12-15

Show video