43:42 Video

Cloudy with a Chance of Pure Storage: Pure's Cloud Focus Yesterday, Today, and Tomorrow

This cloud strategy discussion will cover user requirements and desired outcomes when evaluating cloud related challenges, including agility, cost, portability, and availability concerns. Our Chief Technology Officer Rob Lee and Principal Technologist Andrew Miller will discuss how we solve cloud challenges in five key areas.
Click to View Transcript
00:01
Mm Hello welcome to cloudy with a chance of pure storage. Here's cloud focus yesterday today and tomorrow. Thanks so much for joining us at this deep dive session in pure accelerate tech fest 2022 I think I said all the words right just in order. I'm Andrew miller, I'm joined today by Pure is chief technology officer rob lee.
00:27
Well I'm a Principal technology here Pure and actually linked to rob, you know, some dotted lines, offices, cto etcetera. I'm really happy to have rob with us here today to actually walk through that strategy rob. Do you mind briefly introducing yourself or or actually however long you would like, you know, it doesn't have to be brief.
00:43
Absolutely thanks for having me Andrew. Um it's great to be here today, you know as Andrew said I'm the CTO here here. I've been with Pure almost nine years now coming up this october and in my role I focus on a number of things, looking at our technology and product strategy but also spending a lot of time looking at, you know, where we go next. And a lot of what is capturing our focus areas
01:04
these days is hey how do we help customers uh you know, embed our solutions, our products into their I. T. Cloud strategies and where are they going next with cloud native applications. So really excited to be here today to talk about some of the details. I mean even at a personal level, some of my joint here a couple of years back was to be
01:23
honest around what we were doing in the cloud, I can't hide some of my infrastructure routes, you know, they show I've lost my hair, but the infrastructure roots are still there. But what we've done in the cloud is often more than customers might expect before we get to that though. Wanted to start rob if you don't mind with a little bit of what you were hearing, what you're hearing from customers,
01:42
Obviously we've group them into themes a little bit. People can see, you can see on the slide, but do you mind kind of unpacking some of the challenges that you're hearing from our customers around this area? Yeah, absolutely. And you know, I think in some ways like many uh heights technologies, if you will the requirements and desires and challenges almost
02:03
go hand in hand. Right. And you've done a nice job of grouping them out here, you know, when I talk to customers and, and we dig into, hey, so what is driving into the cloud, what are you looking to get out of the cloud? Um, you know, you have people looking to get agility right? Hey, you know, I, I believe this is a tool that's gonna let me move faster save on costs.
02:21
Hey, you know, these cloud service providers are running at massive scale, they can do it more cost effectively and I have to go build out and manage a bunch of data centers, I can save on cost, portability. Um, you know, hey, this, this is going to give me a lot of access to be able to run different environments, different toolsets and, and you know, I don't have to go and manage all of that availability,
02:41
you know, and simplicity. Right? So these are all the things that I think uh in people's minds when they first go to the cloud. On the other hand, you know, I think what we see is as folks get a little bit of run time, a little bit of stick time, if you will, um each of these presents their own challenges, right? If we look at um you know, if you look at
03:00
agility for example, um you know, I think one of the headwinds that a lot of customers faces, a realization that hey, the skill sets perhaps and the tool sets and the processes and how you're used to doing things in an on prem infrastructure based world maybe doesn't necessarily translate into the cloud. Maybe, you know, there's some re skilling require, maybe there's some re architecture required and so,
03:22
you know, agility kind of runs um well, the promise of agility sometimes run squarely into boy, we've really gotta slowdown retool in order to speed up and, and that's a challenge, um cost, I think this has been well covered. I think um you know, a lot of people are getting a little bit surprised as the bills stack up and you know, some of that is inherent, I would argue and I think some of it also comes
03:43
down to behavior and processing people, right? Um you know, simply picking up and moving to the cloud, but living in an on premise lifestyle, um you know, isn't going to be that cost effective and I think that surprises a lot of people and presents obviously significant challenges. Um you know, portability, I think that's another one,
04:03
right? Um and I think as um you know, especially as people are getting further into consuming all these great services, all of these great um you know, capabilities and databases and what have you of the different cloud environments very quickly, you get, you know, it's very easy to get locked in, right? These are very appealing services to kind of
04:26
build two, but they're also very sticky. And so I think this idea of cloud lock in is really starting to hit people where, you know, years ago, it wasn't on top of people's minds these days, I think people are thinking about it day one, hey, you know, maybe I want some of these cloud resources, but boy, I don't want to, I don't get myself locked into that corner.
04:44
Um and then you know, everything I've described here, right? There's, there's a lot of promise, there's a lot of headwinds and I think a lot of these headwinds and balancing those creates challenges in the simplicity realm, right? So none of what I described in balancing, that sounds simple, in fact, it's actually quite complex in a lot of times.
05:02
So Net. Net I think there's a lot of things that people are looking for but each of these presents quite a few uh and varied challenges in each of these areas. I think sometimes here we'll talk about like the cheaper, better, faster pick two out of three in this case we've got these five listed here, it's like two or three, maybe four out of five like you said with simplicity,
05:22
even a little bit of commentary because I'm just remembering so many stories here like the C. T. O. Who said we want to make everything op X. And the technologists are like does that mean we have to move everything somewhere else? Like just just to make it all affects. Um as well as that when you when you mentioned about portability,
05:37
even just the idea of data gravity. I mean it always makes me feel smart when I say data gravity, but just a lot of data and the speed of light is I think still intractable. I don't think we have anything on the road map for you know, futures to help with that stuff. And then even last just it is about people like you mentioned people process and technology. That's the I think the first Venn diagram I
05:55
ever saw that overlapping centerpiece in the middle is what makes it hard. Yeah. Well and I think the other thing is all of these challenges are somewhat insidious. They don't, they're not obvious day one and it's only until you start getting a, you know, a couple iterations through that you realize, hey,
06:11
there's real challenges that lay here because I, I often have thought of my career, I would like to learn from other people's, other people's pain, I don't know if that's too harsh to say, but it always, you should always start to make new mistakes, mistakes. Yes, with that then, so this is some of the landscape, that's what we're seeing is, we're talking with customers and it is truly like as
06:31
we're coming in there asking us for help in these areas. So this session is about how pure can help. Right? You're attending Pure Accelerate Tech Fest 2022. So what we want to do is actually walk through almost kind of five different ways or areas that pure can help. The first is going to be cloud capabilities on
06:48
premises in ways that you might not expect. Uh then go into kind of a second into consumption based solutions topics on permanent cloud third, enabling traditional apps to have enterprise data services in the cloud. 4th may be called a bridge to the cloud with the cloud adjacent solutions and then fifth, next gen application capabilities, kubernetes microservices, 12 factor app development,
07:11
all the, all the good buzzwords, their their legitimate too, but if we use them the wrong way, they're buzzwords. So before we even dive in here, I want to make sure to highlight to everybody what a lot of what you're going to be walking through I believe is looking at the pure portfolio in what is available today because you look at it in multiple ways,
07:28
both being very clear about what's available right now and then of course you do you think a lot about where it's going and where you might change it but want to highlight the folks that a lot of what we're going to cover today is available right now from pure though, I have a feeling you'll be sprinkling in some future looking items too. So with that let's dive in, I think two, section number one,
07:48
which is cloud capabilities on premises, there's some fun stuff here for me, even as I've dug into the history of Pure and some of the architectural and design principles, I'll be good and sit on my hands and not go super super deep, they're technically, but if if customers want cloud capabilities in their own data centers, what do you what have you seen as the challenges that they're facing and keeping them
08:11
from achieving some of that on prem Yeah, yeah, absolutely. Um Well so first I think it's important to define what we mean by cloud capabilities, right? I think if you ask five customers, you might get six different answers. Um you know, I'll take my spin on it, which is um you know, I think the top attributes people are looking for our agility,
08:32
flexibility, uh, the ability to, to start small and grow in different ways without having to plan years in advance and you know, and this idea that um, you know, as much as possible, uh, you know, the technology just evolves and gets better and better over time that you don't have to spend so much time thinking about how the things are built and delivered and you can spend more time thinking about,
08:54
hey, what you can build around it, right? So, so if we look at those main attributes right? I think, I think it then becomes apparent with a lot of the challenges that customers face today are right. If you look at the state of the rest of the industry in terms of where traditional and prime infrastructure has arisen arisen, I'm not sure that's a word has,
09:16
has risen from, you know, a lot of these, a lot of these systems, you know, really designed in an era where if you had an application environment, you have a thesis of how that behaved, you had a thesis of how big it was going to grow, how fast and so on so forth, especially crafted a little, you know, island around it and you know, that was probably a fairly involved exercise,
09:39
you put work into it. It took time and once you got it working boy you don't touch it right. If it ain't broke, you don't touch it and you know, that worked well for a number of years or perhaps even decades. Um but if we think about this in the context of cloud capabilities and agility, it's kind of the antithesis,
09:57
right? Um and so I think that, you know, this idea that hey, I want to get agility, flexibility, evolve ability and continuous improvement and build this on top of legacy infrastructure that is not designed to scale, not designed to grow, not designed to flex, not designed to be multi purposed and recombined in different ways. Um it's worse than a square peg in a round hole,
10:22
but I can't come up with a better analogy at the moment and it fits eventually. You just hammer it hard enough and it fits. So I think, I think that's one of the biggest barriers, you know, I think the other, the other thing that goes hand in hand with that is just the whole management experience around uh the traditional on prem infrastructure,
10:43
right? Um you know, and even at the simplest level, um if I go to a traditional cloud service, right? Um I care deeply about the attributes of the service that I'm provisioning and paying for, I care deeply that the eye apps that I'm paying for, I'm getting, I care deeply that the um, you know, the amount of deer and that I'm getting in a virtual
11:04
instance that you know, that I actually get that delivered. Um I care deeply that there's great hardware delivering that. I don't want to know its name, I don't want to know where it sits. Um, you know, I want a little bit of an abstraction away from having to manage those details. And I think that historically, you know,
11:22
if you look at the legacy on prem solutions, they just haven't been able to bridge that gap. And I think that if we turn our attention to what we're doing at pure, well I think bucket number one, right just in terms of addressing agility, addressing, you know, future proofing and the ability to grow and scale and evolve in different ways. This has been what we've been at since day one.
11:44
This has been in our D. N. A since day one of all of our product lines And you know, hopefully with today's announcements in Flash Blade. S you can see were taken, you know, we're taking those concepts to the next level, you know, in Flash Blade. And then I think on the management side, I think this is where um you know,
12:02
with pure fusion, what we're doing with the fleet wide management with bringing a lot of the cloud like management model to our arrays. I think we're taking a huge step forward in bringing that operating model if you will um, out of the cloud and bring that into customers on prem environments. I thought sometimes of it, it's not making our for forcing our customers to make payments on
12:27
our technical debt that we just happened to put in their data center and, and, and even sometimes a lot of it is that I think we've heard for a long time. See IOS developers, there's a lot of things they would rather not have to care about and but it takes actually really hard work, architectural work under the covers to make it possible not to have to care about that. That's even not even wanting to do stuff like
12:46
waterfall versus agile and, you know, cattle and cattle versus sheep. Um, I'm losing my analogies here, cattle versus pets. There we go. You know, I didn't quite click Okay, anything else that you want to add from a kind of how it pure is doing in this area or historical where uh, I don't think we're doing like the crazy hand signals here,
13:04
but I don't want to go into number two and leave anything else out there that you want to toss in. Um, No, I mean, I think, I think this is, you know, the, the areas that we've talked about, I think are the key focus areas, Right? Hey, how do we, how do we really bring the promise of agility and flexibility and future proofing?
13:21
Um, you know, to traditional on prem environments, how do we bring more of the cloud operating model which again unlocks agility, right? The less you have to focus on the run and operate, the more you get to focus on building things around, right. Um, and I think we're making huge strides in both of those areas and um, you know, I think we're, we're,
13:42
you know, in one, um, we've been about since day one with Evergreen and, and certainly now the evolution of Evergreen and the new announcements within the evergreen family. Um, and then with pure fusion super excited to see where we go with that and kind of what's coming next, but we're really just getting started. So, um,
13:59
you know, if you haven't, if you haven't dived into the details in either of these areas that would highly encourage folks in the audience to take a look, you know, we've got some sessions on those, we've had a great launch event last year on pure fusion. There's some fun demos, there's new information there too as well. Second, a little bit of possibly,
14:18
uh, you know, kind of step into the side, although he said it's number two. So consumption based solutions, if you're listening to this, you're probably reading this and thinking, hmm, I've done some stuff in the past that was like dressed up leases, that kind of thing, that there's been a lot of talk in this space because there continues to be desired, whether it's at a CFO level or aligning spending to
14:39
actual, you know, revenue generating activities, you know, we don't just think a bunch of complex and so there continues to be this desire around apex on prem but also in the cloud obviously. So do you mind kind of unpacking that as well? You know, the challenges, you see customers facing their, how, what's keeping them from achieving their
14:54
goals and how we're leaning in and dare I say, some unique ways. Yeah, absolutely. So you know, I think first and foremost if we set the cfos aside for a second, I think very few customers are actually driven our goal to get from where they are to apex, right? Apex. I think Capex versus apex is, it's a side effect of what they're trying to achieve, but I don't think it is the,
15:17
The one and 1 and only goal. Um, I think what customers are really looking to achieve and, and um, and I see this pretty universally in all the clients that I talked to is additional flexibility and future proofing, right? And this kind of again, not to beat a dead horse, not to, not to, you know, banging this drum too much, but you know,
15:37
similar to topic number one. Um, you know, if we think about the more traditional acquisition model, the more traditional product experience, um, you know, you buy an asset before you buy that asset, you have some concept of what it's being used for what that application, where the environment is gonna need today, what it's going to need in a year,
15:58
what it's going to need in a couple of years, you have to do some planning, you have to do some forecasting. Um, you have to basically predict the future and then back that into today and say, okay, well what do you want to, what do I want a provision today for the future and um, you know, and then go from there and then you have to map that into,
16:16
um, you know, uh, product Scuse, you have to map that into, hey, what is it going to take to go and acquire run and operate that? Um, and I think that there's a couple of significant advances that we're delivering on top of that experience, right? Like that makes a lot of sense for, for people and we make that experience great with Evergreen forever in that.
16:37
Hey, we can take you through that evolution, we can take you through changes over those years. Um, but I think the last couple of calendar years have shown us that the ability to predict two or three years out is, you know, questionable in a lot of cases. And so, you know, I think what customers really want in a true consumption based, true outcome oriented based offering is protection from the
17:03
unknown. They want investment protection. They want flexibility to evolve and you know, when we look at things like what we're delivering with evergreen one, when we look at things like what we're delivering with everything and flex. It really ultimately comes down to the flexibility.
17:19
We're giving customers to not have to plan for perfection three years down the line, four years down the line and be able to burst as they go right. And yes, as you kind of unpack that, then you get into, okay, consumption versus upfront. You, you know, you get into um you know, all the way down in today's complex topics.
17:41
But I think starting at the financial treatment and then working your way back um you know, it's only inevitable that folks that look at the offerings that way end up with a dressed up financial services, professional services type of thing around us. You know, overly bloated menu of products dressed up around T. S. And C's to look like a service.
18:04
Um You know, whereas I think when we look at what customers need around consumption based outcome or services, it really starts with technology. It starts with product capabilities and customer needs and it really is an engineering lead exercise that we can then build a business model around. I mean I think of, I love the planning for the unknown or I'm changing the words of that.
18:26
Planning for the unknown. And I'm even gonna go pull out some of my old economics classes and talk about like opportunity cost and optionality because you bake into that multi year thing and you can't be agile. It goes back to our previous slide and even I'm remembering sometimes as a, as an sc, when I was a partner. I say for all,
18:42
you know, I'd go in with customers where I had better relationships and we'd be very real. Like we're not going to get it just right. Do you want to go for too big or too small? Let's decide which way we want to go because we know in a couple of years, if we can get it just right, we should be like playing the stock market not doing whatever else we're doing. But from from an alignment to the future
18:58
standpoint. Being able to shorten that runway of how far you have to have visibility. So critical. Exactly. Exactly. And you know, and it's not even, it's not even in the aggregate totality. Right. If we look at the newly announced ever been flex for example, um, it might just even be within, hey,
19:15
um, you know, within this site, I want to run the hardware, You know, I want to run and operate it. I maybe have to do that for contractual reasons. Maybe I have to do it for regulatory financial compliance reasons we get no problem. But hey, can we get additional flexibility into user word is a great word optionality. Can we, can we get some additional optionality within that site to say,
19:35
hey, if my thesis on day one for where this capacity was needed across these different arrays changes in, you know, day 180 changes and you know, after a year, um can I have the flexibility to just kind of move that capacity around and and redeploy it, We know the product, we built the product to support that. We've built the technology to support it. And the nice thing with evergreen flex now is
19:58
we've got the business model, we've got the consumption based um subscription to go along with it Both ways. I think that takes us into number three, which is like it says on like it says they're around enabling traditional applications with enterprise data services in the cloud. I realize that's a mouthful but but this goes to some of the idea that it's easy and actually
20:22
this is even a personal thing. Like we all take things for granted until the power goes out until the water stops right. Like we think of certain things as utilities almost there's like a great thing about, you know, I. T utility if you go back, I'm forgetting the author's name, but even when it comes to on prem capabilities that we've had sometimes as far as data
20:38
services or even storage capabilities. Often when we move into the cloud, sometimes that's some of those capabilities aren't they're not just because it's like it's good or bad. It was just came from a different design center. So I'm thinking about now about talking about customers who have existing applications,
20:55
they're trying to figure out ways to put these into the cloud. Their thinking about enterprise data services and the richness of stuff available, especially if you're a pure customer. Hopefully you are. What are the challenges you see customers facing is they're going through that thought process and maybe, you know, the bumps that they run into along the road and then of course
21:12
how we help? Yeah, absolutely. Um, so I think we could probably do an entire section on this, but we've got limited time and attention. Um, you know, I would call that two major challenges or categories of challenges and I think one you touched on, which is, hey, if I've got traditional applications,
21:30
they were built with the thesis of the supporting infrastructure is we look to the cloud, if that infrastructure was built around a different thesis of the applications that we run on top of it. Um, you know, these two things may not be compatible cattle versus pets. For example, traditional apps expect the infrastructure they're running on to be highly curated,
21:47
highly looked after pets. If you're trying to then run that in a more cattle like environment, how do you deal with these unknowns and, and and unexpected consequences. I think that's number one. Um, I think number two is also that, you know, I think if we step back and and those of us that have been in the industry for a while and have had multiple jobs and exposures to
22:09
different environments, I think we'd all agree that different enterprise environments all look a little bit different. Right? The network configurations are all different, um, you know, there's a lot of differences as you move from enterprise data center to data center. Uh if you look at the cloud service provider environments, um,
22:28
those tend to be highly opinionated environments in terms of how networking compute storage are set up. They tend to be highly opinionated environments around things like security, things like VPC, definitions and so on and so forth. And oftentimes again, you have a little bit of an incompatibility there of hey, if I'm used to running and operating in this environment, how do I then move and and clamped
22:51
to somebody else's highly opinionated environment building blocks. So I think those are the two major challenges, you know, that we see um, and to simplify it a little bit, if we can edit out, it comes down to a mismatch of expectations of the application and capabilities in the environment. Right. And so, you know, are, you know,
23:12
here appear our strategy, our view, our our place in the world and trying to solve this mismatch is to really be that translation layer, if you will Right, we want to present the same um, services, the same capabilities. The same feature sets that customers expect customers rely on us for customers are used to us for in the traditional enterprise environment but then repurposed re platform and
23:36
re optimized for the cloud service environments, We want to take on that complexity of, of managing the differences in these environments. So when you look at things like cloud block store, for example, you look at that offering the northbound interface right to the client applications to the customers that looks exactly like purity running on our flash arrays.
23:56
The southbound interface, if you will. Right. What that's actually running on is a completely different beast, right? It's just like another hardware version, another array version that we have. It's just happens to be called AWS or it happens to be called Azure and that's the layer of secret sauce where we're really bridging a lot of those gaps were providing the durability, the availability,
24:17
um, you know, the rich data services that are lacking in the cloud service provider primitives. And I think ultimately for a lot of the use cases that customers are looking to achieve as they think about traditional apps in the cloud, you really need that translation layer. I'll give you an example right, A very common one that we hear is customers saying,
24:39
hey, I'm gonna run my production, you know, data center, I'm gonna run my environment on, on flash arrays. That's not gonna change, We love the performance levels, attributes, but gee, it would be nice to be able to do D R to the cloud, right? I don't have to fail over that often.
24:53
Hopefully. And if I do, you know, it cost really isn't that much of an issue. Um, I just need to get there and get back well in order to do that you've got to be 100% sure that you know the thing you're failing over to write the storage array if you will that you're failing over to behaves exactly the same as the one in your production environment or else Exactly well and that's exactly it.
25:18
Right the last point in time you want to discover, oh there's a caveat oh this little feature that you depend on you know significantly behaves differently in the cloud environment is in a disastrous scenario. Um And so I think I think to net it out I think the real challenges customers face and it comes in different flavors is a mismatch between application requirements and expectations and what's available in the cloud.
25:45
And our job with whether it's cloud block, store whether it's sport works, whether it's you know, things that come down the line is to bridge that gap and provide that environment agnostic experience to applications and administrators. This even for me it takes me back a little bit too like systems theory and the law of unintended consequences. We built really complex systems inside the data
26:06
centers and sometimes we don't even know what assumptions underlying architectural assumptions don't apply anymore. And then even going all the way back. I love that you said you know thesis thesis thesis that I'm thinking of like thesis antithesis synthesis Hegelian dialectic we can insert a little bit of philosophy into I. T. At times,
26:23
you know kind of thing. But often I find conversations like this go one of two ways either customers have found some of these challenges and they're looking for help or sometimes I'm having to tell them things that they don't really want to hear yet, but I'm the bearer of bad news and you try and do that as diplomatically as you can because they were there to help but they almost don't want to hear it or they're not ready to hear it
26:42
yet. Okay with that let's shift into we're almost coming down the home stretch so forth. So you might be thinking like okay so bridge to the cloud with cloud adjacent solutions. What exactly does that mean? So so bridge feels kind of kind of positive like it's getting us from here to there, Cloud adjacent, it's sitting kind of close but
27:05
there's a we've actually been doing a lot of innovation in this space and I think it ties really well to number three of people in the industry seeing that they need some new options and some new capabilities which we even ran into some of those challenges earlier with even maybe pure one. I'm I'm leading the witness here. I know but I'll at this point I'll shut up. And do you mind actually kind of walking
27:25
through some of what bridge the cloud with cloud adjacent solutions means and how we help. Yeah, absolutely. I see where we're tailing into the buzzword section of the session here. This is good. Um no, so I think, I think um you know what, when I think about the demand for cloud adjacent solutions, I think this is actually a good indicator of the nuance that people have
27:47
developed around cloud, right? The 5, 10 years ago, five years ago, um you know, the conversation was on premier cloud, I'm gonna stay here, I'm gonna move all over there. It was very binary. Right? And I think the nuance that people are developing is hey,
28:02
we've got different tools for different purposes. They're good at different things and you know, depending on my needs, I'm gonna, you know, I've got a hammer, I've got screwdrivers, I got socket wrenches, right. You know, they all serve different purposes. And when we think about the demand for cloud adjacent solutions,
28:17
I think this is a good combination. These are a set of use cases where customers are seeing the benefit of combining the power of different environments. Um, we see, you know, cloud adjacent demands and I would say in a couple different places. Right. So I think one is um, you know, areas where I would say customers want control of the infrastructure choices,
28:42
maybe some of the run and operate pieces getting back to our previous part of the previous part of our discussion, um the, you know, a more opinionated to them stack of of how pieces are put together, but maybe they want to get out of the brick and mortar running a data center business. Right. How do we bridge that gap? Um, I think there's another class of demand where we see customers saying,
29:03
look, I have a lot of data that's being generated or being stored in an area or whether it's on prem physically whether it's in a cola, but in my kind of direct control it needs to stay there. But boy could I use cloud based resources, whether it's services, whether it's compute to be able to interact with that data, wouldn't it be nice if I could combine those two things together.
29:25
And so I think we see both classes of demand. You know, I think that we, one of the ways that we're serving both of these classes of demand is really looking to partner with key service providers, managed service providers that have such as Equinix with which we have direct partnership around their bare metal as a service product to really support this entire spectrum of needs.
29:50
Right? So if you look at bare metal as a service by Equinix that's powered by here, at the service under the covers. So customers that sign up for that service, get the benefit of offloading some of the run operate brick mortar aspects of data center kind of management to equinox through that service, but they get the benefit and they still get the benefits of best of breed pure
30:10
technology um, without having to go through all the pain we just talked about in terms of re platform into the cloud. Um, the other thing that uh, you know, cloud adjacent service providers such as Equinix provide is very fast and thick data pipes into the cloud. And so when we think about areas of solutions like technical computing media is a great
30:34
example. Chip design. You know, we've got a partnership with Azure and economics to support customers in that space that need to want to and maybe for data gravity reasons have to hold their data on our physical arrays. But by pairing them in Equinox to Azure compute can get far faster, far more compute resources applied to that data and get those simulation
30:59
jobs get those chip design jobs completed much faster. Another example of this that we recently announced is our partnership with Snowflake. Right where again, we're pairing the power of flash blade, the object storage fast object storage platform. We built their design for analytics with one of the leading cloud data warehouses in the space snowflakes so that customers can hold their
31:22
data on our arrays. Perhaps even share that data with other on prem and analytics tools, but then also apply the power of snowflakes data warehousing capabilities to work on that data in the same place. So net. Net, you know, I think we're seeing the beginnings of a growing space of cloud adjacent solutions in demand and I think it's a great
31:44
indicator that customers are getting a lot more sophisticated and a lot more nuanced about putting together the pieces of, hey, I've got tools on prem, I've got tools in the cloud, I've got tools maybe in cloud based services, how do I choose the best of breed solutions and our strategy here at pure is to really enable the connection of those best of breed services.
32:05
I love you kind of, we said data gravity all the way back at the beginning but I think what you were just talking about is away almost to make data gravity work for you instead of against you potentially kind of thing. I also think there of time to value, you know that we actually often for a business standpoint we want to get value more quickly and if we can avoid some of the physical stand
32:22
up pieces, just shipping gear takes time standing up networks, links takes time. We partner with companies that are really good at doing this already have it already done. And then of course the go to the cloud, go west young man, you know, kind of thing. Like if all you have is a hammer,
32:33
everything looks like it now I think 5, 10 years ago that was more of what was going on just to the cloud around prim and now we've realized eventually you can, you can hammer a screw or anything else in with a hammer, but eventually start to look around for something more than a hammer. You need a little bit more nuance and it feels like an indication of the this space getting
32:50
more mature. Just as an industry, we are absolutely okay, we're firmly in the buzzword territory now. So next gen applications capabilities and okay, so let's unpack it a little bit. If we're doing next in application development is probably about microservices for rapid scaling and everything be ephemeral and disposable.
33:09
Now we're firmly into pets versus cattle. Of course, then that probably means you might care about containers, you want lots of them. So now we care about kubernetes for automating all this stuff and we have good application design principles, liked olfactory app development. Now this is an area where if you're listening
33:23
in, you might be thinking like why does pure have a section about this? Right? What is, what is pure doing here? So I think rob, I'm not going to steal any of your thunder here. Do you mind just launching into this? Yeah, absolutely. So I think you, you had a nice little, I would say walkthrough buzzword
33:42
territory from the bottom of the technology stack. Let me try to come at this from the top down, right? When we think about next generation applications, um, you know, I always look at one of the problems that people are trying to solve and what are the major trends and really secular trends and shifts the industry and my step back from it,
34:00
I think the problems that, that people are trying to solve is it really comes down to 22 things. Right? One is they want more portability, They want they want more portability as you're building an application these days. You want that application, you want flexibility and optionality to be able to deploy that application on one
34:19
infrastructure today and move it to another one tomorrow and Bridget to a third area um you know next year and not have to redesign that application from scratch. Right. So I think that's you know, that was an afterthought for for many years. I think it's a primary thought in application design today. So that's kind of driver number one,
34:37
I think driver number two, is that um the software ecosystem around us is just so much richer than it was 10 or 15 years ago, 10 or 15 years ago or 20 years ago, if you were building a large software stack, you kind of had to build it from the like assemble the molecule at the molecular level. Right? You had to build, you know, um you had to build everything from,
34:57
you know, maybe a custom data layout, iO schedule or a thread schedule, er you know, distributed configuration manager and all kinds of stuff. Um these days when people build applications, right? They want to be able to go and reach into the rich ecosystem of software tools, they want to be able to reach into open source and get the benefit of that to accelerate them
35:18
and say, look, you solve that based problem, I'm going to focus on, you know, my value add, why I'm building a software application and so if you look at those two drivers, right, I want portability, I want access to the open source ecosystem. Well then naturally that leads me down to things like containers, things like kubernetes, Right? These technologies are very,
35:38
very lend themselves very, very well to that. They allow me the ability to virtualize small pieces of my application, they allow me the ability to independently manage and version control the 15 different tools I'm building on, Right? Um and so that naturally leads into, okay, well, I'm going to design my application around a, you know, potpourri of different application tools or microservices,
36:05
if you will, I'm going to use containers to virtualize that. I'm gonna use kubernetes to orchestrate all of that. All right, So now, once we're there, um what are the challenges that people face? Well, you know, I think in a non technical way, the journey that I subscribed is kind of like, you know, building up a car from uh from a kit car, part,
36:27
like from parts, right? Um and so you can get there and you and you will, but not every car is gonna look the same and in fact they're probably all gonna look a little bit different because the pieces are going to get assembled very differently, it might be a lot of fun, but if you're taking your wife somewhere or doing anything, you're not gonna do it for a while. It's gonna take you a long time together.
36:45
You're not going to do it for a while and um, you know, and then think about it at scale, right? Yeah. Imagine if you're, you know, a rental or I'm going to abuse the analogy now, but you imagine your rental car company, right? Um, organization, right? You want to manage the same thing over and over
37:01
again the same way if every one of your cars looks different, like that's going to be a big problem. Right? And so I think there's a couple issues that, that companies, that developers that, you know, I. T. Managers run into in this space, right? Number one is, um, it's new and people are
37:17
still learning, so there's a feeling of skilling gap, um, and it takes a long time to get started. Number two is, there is a lot of changes, a lot of different moving parts, right? It's not just, you know, it's not just, hey, I'm gonna go get oracle version, you know, nine, I'm dating myself now and I'm going to
37:36
configure and deploy it. It's, I'm gonna, you know, I'm going to version control and manage 30 different, you know, components in my system and if there's a security problem in one of these applications, I have to go and chase that down and management upgraded, know, if that brings a different dependency, I'm gonna manage that.
37:51
Um then there's also the fact that, you know, kind of kind of coming back to the promise and the reality of things. Um, the open source ecosystem offers the promise of access to all this great software and all these great tools. People love talking about open source developers. When was the last time you heard somebody talking about open source QA person?
38:09
Right. So there's the reality that, you know, things change quickly because bugs appear they get fixed, but all these things lead to a lot of change. And so Net net, right. Your challenges are complexity management, uh, you know, ability and speed of getting started and then on top of it all,
38:28
you know, making this stuff rock solid, Right. Um, if you want to actually build something for real and get out of the uh, toy kind of science project phase, you need all the enterprise capabilities. You need B C D are you need security, you need all that stuff too. Right. It's pretty important. And some of that stuff for the open source communities, I'm even thinking of like the
38:46
cathedral and the bazaar by eric Raymond way back in the day, there's a lot of fun stuff to do. People, white people are motivated to do it, but some of the stuff you just described is critical. It's not necessarily fun to build necessarily depending on who's out there in the open source communities. Yes, it's a often been left as an exercise to
39:03
the reader if you will. Um and so if you think about and again, I've got a little bit long in describing the background, the context and the drivers, but if you look at that context, then um what we're trying to do a pure, what we're trying to do with poor Works, why we're so invested in the spaces. Well, a you know, we firmly believe that this is a secular shift,
39:23
right? There is, you know, software is being built this way you go hire somebody out of school, this is how they build software, right? So this is the new tool set. Um and there's a lot of need to solve these problems. If you look at what we're doing with pork works, we're addressing each one of these head on,
39:36
right. Um if you look at poor works data services, right? It's primary goal what we're doing with Poor works data services is bridging the gap between, you know, the promise of open source and the realities of it and really making it tenable saying, hey, mr developer, Mr developer, you want to build on cassandra or Mongo or Postgres great
39:58
um will help you do that without you having to become a deep expert in how to configure a chafer three nodes of cassandra, Like we've got you right on the flip side, we're also helping the I. T. Managers with the kind of run and operate pieces of that because one of the problems I often hear is hey I've got five or six teams building next gen apps on containers, they build something,
40:23
they toss it over the fence. I've got, you know, everybody's using cassandra but I've got six different configurations for it. Do I know like is this configured differently than this because there's a reason and somebody thought this through or is it that these two developers just went to two different stack. Overflow pages. Right.
40:40
And so by bringing bringing curation, bringing control into the system, we can help administrators bring a lot of the compliance and just kind of, I would say enterprise hygiene back to next gen applications. Of course none of this would work if it weren't backed by the other layers of pork works, which is the rich kind of BCD our capabilities, whether it's back up,
41:02
whether it's D R migration that we have a PX backup and the core underlying kind of enterprise grade but container native storage layer in PX enterprise PX store. Um and so, you know, net net right. What we're trying to do is bridge that gap between the promise and the need to run fast in new application design and development but also the need to have enterprise hardened attributes and run and
41:29
operate hygiene around these application environments and for those so many different themes that I love like a level of risk versus reward. We have to run fast but without with with with actually mitigating the risk even his I. T. Teams sometimes. Yeah, I mean it comes down to uh you know, I think famously said run fast and break things.
41:49
Our goal here is to let you run fast without breaking, get the first half of the second half exactly. I don't think we took the mean any good analogy isn't good until you stretch it past the breaking point, but I don't think we took the kick card car in the garage to a rental car company. We didn't start it quite to the point of breaking, you know,
42:05
so but hopefully that helps people kind of understand a little bit of what we're doing here and even I love how you highlighted both the infrastructure next gen infrastructure level, but even all the way out to, you know, delivering databases that that sport works data services and continuing to push forward on that vision with that I think we're actually through overall what we're planning to go through what I love and some of what we walked
42:26
through is there were relatively few product names and I was even kind of playing a little bit of that home game as you were talking about because like you're the CTO right and the goal is to focus on the product categories and the capabilities. You heard some product names along the way, Right? To illustrate that were we legitimately play in these spaces and we do things where we can help
42:45
you if you want to find out more about this, the various products and deeper dives, please make sure to go what it says, they're pure storage dot com slash cloud, you'll see a whole walkthrough of a bunch of the different capabilities and products that we talked about with that rob. Thanks so much for joining today for making the time. Do you have any last words to any last thoughts
43:04
to bring us home? No, I was going to say thank you for having me and uh, and for those of you in the audience. Thanks for tuning in. Um, I hope this was informative and, and as Andrew said, I focused a little bit more on the underlying technology, the trends, the drivers are investment thesis, but I highly encourage you,
43:22
uh, you know, to, to go follow up, learn more about our cloud solutions. We flashed the U. R. L. Earlier uh, and with that, Thanks for having me. Awesome. Thanks rob. Thanks everyone for joining us. Have a great Accelerate Tech Fest 2022
  • Bare Metal as a Service
  • Evergreen//One
  • Video
  • Pure//Accelerate

Pure has embraced a cloud operating model on-premises and in the cloud since its origin to solve our user’s biggest business challenges. Our Chief Technology Officer Rob Lee and Principal Technologist Andrew Miller will discuss how we solve cloud challenges in five key areas.

  1. Cloudy capabilities in on-prem storage
  2. Evergreen//One: a true consumption-based offering providing OpEx on-prem and in the cloud
  3. Enable traditional apps with enterprise data services in the cloud
  4. Bridge to the cloud with cloud adjacent solutions and Bare Metal as-a-Service
  5. Next-Gen app capabilities (Kubernetes, microservices, etc.). Twelve-factor app time!

View All Pure//Accelerate Sessions On Demand

Watch Now
Continue Watching
We hope you found this preview valuable. To continue watching this video please provide your information below.
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
CONTACT US
Meet with an Expert

Let’s talk. Book a 1:1 meeting with one of our experts to discuss your specific needs.

Questions, Comments?

Have a question or comment about Pure products or certifications?  We’re here to help.

Schedule a Demo

Schedule a live demo and see for yourself how Pure can help transform your data into powerful outcomes. 

Call us: 833-371-7873

Mediapr@purestorage.com

 

Pure Storage HQ

650 Castro St #400

Mountain View, CA 94041

800-379-7873 (general info)

info@purestorage.com

CLOSE
Your Browser Is No Longer Supported!

Older browsers often represent security risks. In order to deliver the best possible experience when using our site, please update to any of these latest browsers.