00:01
Right. Good afternoon everyone. And thank you so much for being here in this session. I know we probably are the last thing between you and whatever your evening plans are. I was going to go with beer, but then there were some alternate plans as well. So let's jump right quickly in.
00:21
Uh we are here to talk about why pr for your file and object storage system. We'll touch on a few common topics that we hear from customers around their primary challenges about their file and object storage and how we as pr look at this and why flash blade blades approach to these specific challenges. So, I'm BJ Misra, director of product marketing here at pure storage and I have with me
00:49
of uh Andy, uh from the product management team here at Flash and he like I promised will answer all of your questions that you have at the end. A quick show of hands. How many of the people here are existing pure customers? Keep your hands up if you also are a Flash Blade customer? Ok, I'll ask a question, how many of the
01:17
separate question you can put your hands down? Uh how many of you are looking at a file and object storage system in the next 12 to 18 months. Hopefully some of this discussion will help you in like, you know, evaluating your choices. And like if you have any follow on questions, we'll absolutely have uh time at the end.
01:40
So let's get started if we, when we spoke to the customers about the five big areas, that was the major challenge for them when it came to file and object storage. These were the five areas that came up. The first one would be rather unsurprising was the data growth. But what was a little bit of a surprise in there was along with the data growth.
02:06
It was the application sprawl associated with that just exponential growth and application. The second area which they mentioned was a big challenge is the complexity of the architecture, how difficult it is to manage and like, you know, deploying it scaling it to meet their requirements. The third area was the power and space requirements associated with the product.
02:32
The fourth area, which was an interesting way when they framed it in different conversation is just the rigidity of the purchase cycle. The disruptive operates every 3 to 5 years, having a situation where they have to refresh the storage environment and the entire experience that they go through. And the last point was around their demanding and ever changing business requirements.
02:59
All of the evolution and growth in data and application is also leading to an increased demand on the storage environment. And the slas associated with with storage have are evolving over time and they want their storage system to be able to adapt to it during this session. We'll touch on all five of these areas in a little more detail.
03:26
Let's start with the data growth challenge. Now, the fact that data is growing isn't really a surprise to anyone but when we speak to customers, one of the things that comes up is like, you know, 80% of that data is unstructured. Now that for some customers, they, they're a little bit surprised when we say that we tried
03:52
to look at some of the research available in there. The on prem file and object storage by 2025 is expected to be 100 and six exabytes. And depending on some of the analysts projection unstructured data is expected to grow by 10 x by 2030. The big takeaway from this is this growth in unstructured data and application is expected
04:20
to continue unabated for the near future. But the data growth doesn't tell you the whole picture. The reason the data is growing is because of the emergence of a lot of technologies from analytics from IOT artificial intelligence, machine learning and and so on. And it's impacting a variety of industries from financial services and where it's like,
04:47
you know, being used in fraud detection from like manufacturing and retail where it's used for predictive fulfillment or an automotive where it's like, you know, driving self, self driving car technologies. These are just a few examples but the list is endless. The reality of it is this growth in unstructured data is spanning across use cases
05:12
across industry and vertical. And that is leading to some growth in some new and emerging industry trends. Unstructured data is really tier one storage right now. That wasn't the case probably a decade or so ago, object storage, particularly high performance, reliable, fast object storage is becoming mainstream,
05:39
being used in applications from everything from analytics right down to data protection. I will not spend a lot of time on this. But the fact is if the demand for enterprise A I is all around us, it's mainstream that application probably touches every customer's industry in some way or form. But the net net of it is there is a demand for a simple,
06:09
reliable, efficient storage solution for file and object workloads. Now you might ask that you know all of that might be true. But what's wrong with the current approach? There are plenty of solutions available in the market. Unstructured storage solutions have been here for decades and when they were created,
06:32
they were really good compelling solutions. Now, they were built for a very specific purpose for a very specific performance profile and they really delivered well on that front. But now when we look at the emerging trends in the market and the emerging workload requirements, they really failed to deliver because they are absolutely rigid solutions
06:57
which were great for a specific use case or like one part of a specific section of your workload. It might be a limitation with the protocol. It might be a limitation in terms of the specifics file sizes that they can support. It might be, they might be great at light batch processing, not so good at sequential processing. They might the performance might degrade when
07:20
the file or object count increases beyond a certain level. That's like a typical problem now. And it's across the board. You might have like a typical customer today, might have one or more of these architectures that they are using for various different applications. You might have a database application for your analytics workload.
07:42
You might have like, you know, some sort of, you know NFS workloads on your NAS architecture. You might have an object store for your image uh kind of uh you know health care image requirement, you might have a backup purpose built appliance. But in this, this architecture is what adds to the complexity and data sprawl that's happening in the data center today.
08:07
And that's where pure flash blade stands out. It becomes a single platform that can deliver all of the support for the workload in a very simple seamless manner. It's highly paralleled, it delivers incredible level of scalability on both your performance and capacity and delivers native high performance
08:35
file and object capabilities. And with our Evergreen technology, it's a storage solution that always remains modern and evolves over time to meet your requirement to talk a little bit about how we handle, you know, the data growth and the future challenges. Let me hand it off to Andy. Thanks.
09:00
So Bish mentioned that there's uh you know, most of the world uh is unstructured, mostly unstructured lives on hard drives. Let's just talk about that for a little bit. Um I worked in it, right? You, you buy a an array because you got a project and you got funding. That's the magic words right there. You got funding, you get something in the house, you get it in,
09:19
you got some nice predictable workload on it. Day one and then that's it day two through the rest of its life span. It slowly deviates from that and the beauty of it deviating is you want a system that's flexible, that can deviate as far as possible, right? And unstructured workload really kind of means or unstructured data really means unstructured workloads where you can't really quite predict
09:39
them. I think that's what was trying to get at is we have a very good system here for unpredictable environments because it's very good at scale out. It's very good parallelism. But when we talk about the capacities and kind of the growth in the world, um you know, hard drives are good at sequential work.
09:54
They're not really great at uh you know, at random work. And in general, you end up with a lot of, lot of spindles out there. And so part of the problem we have is that sprawl of the physical hardware. And so as we saw, you know, going back about 10 years ago, right, flash started coming into the environment.
10:11
And so you start putting in a little bit of flash, we start seeing a little bit more flash, a little bit more flash, a little bit more flash. And so the the real problem we're starting to see that, that everybody in the industry is seeing is the flash is growing. Um the capacities of flash are growing much better than the capacities of hard drives are
10:29
growing. And the amount of data that you're being asked to put onto the systems is outgrowing both of those. And so the question really then becomes sort of, hey, we have all of this unstructured data which needs an unstructured workload that we can't predict and we need a system that kind of scales and grows with us.
10:46
And so, you know, now people we look at pure and pure as a flash based company. Um And it's kind of funny because often times people will ask us, you know, we're, we're a flash based company. They always ask us, when are we gonna put hard drives into our environment? And we're like, well, you're not really asking us when you're gonna put hard drives in,
11:00
you're really asking us when is flash gonna get cheaper because hard drives were pretty cheap. And so you could buy a lot of them. But when you look at flash, you always looked at for performance, right? And typically in hard drives, you always measure them in gigabytes a second.
11:14
But that's not really the best thing for flash. I mean, flash can do gigabytes per second. But what's really great at doing is IOP it's really great at doing metadata operations, you know, multiple deletions, all those other things that make that unstructured workload very, very difficult to manage in a hard drive environment.
11:29
And then as you start seeing flash kind of start leaning over towards capacity, now the question on capacity becomes, you know, the question is on capacity, right? Terabytes, but it's not just terabytes, it's also things like the number of objects, number of exports and the number of large name spaces, all of those other things over here that you look at for capacity.
11:46
And so now as flash starts taking over more and more of the capacity you're starting to now look at, hey, what differentiates this in the world and what really differentiates. And the reason why people are looking at it is because of this chart. This is a chart I've been in storage for 15, 20 years. We've all been looking at something like this for a very long time.
12:04
When is that price point where flash is going to overtake disc, I can get rid of the spinning discs. I don't have to worry about those flash is more reliable and I can finally just start to go in there and use solid state in my environment. If you look at this, the crossover point is approximately 2027 that's about three years from now, four years from now. So why are we even here talking,
12:24
talking about this chart today? And the reason is because that is on a, you know, a and SSD basis. Uh If you look behind the scenes of the data and where it came from, they basically go every year, they go shopping and they go pick out a device that's the largest capacity, lowest price and they throw it on there.
12:42
But if you look at what we're talking about here with the 2027 the crossover point, the crossover point is actually today and that's because we can achieve that on a full system basis. And that's the reason why we came out with the flash blade E. And that's the reason why we came out with the flash array E because on a full system basis,
12:58
if you look at the total cost cost ownership of it, we can actually meet the price of hard drives today. And that's the reason why you keep hearing the message everywhere, right? This drives are kind of dead. Let's go look at flash now.
13:10
Um And, and I don't, I don't view it so much as disk drives are dead. Let's look at flash, I view it more as we are finally able to use flash in all the places that we wanted to. And we are now at that price point where that makes sense, disk drives will just naturally go away because of that because flash is better at doing that. And so let's talk a little bit about that
13:30
holistic view of that system. And how do we get to that overall TCO where a flash based system can, can is better than a hard drive based system. Uh It starts with the clean sheet design. Our, our systems were designed with hardware and software put together and if you look outside in the world everywhere, every company always starts being,
13:50
we're going to be software only. We're going to be agnostic to hardware, every single vendor that's ever started like that and has grown, they've all come back to building their own hardware, all of the major cloud providers all have their own specs for their own equipment, all the major vendors have it. Uh If you look at any place where you have any type of tight integration,
14:07
you're always going to have that and that helps out with qualifications. It helps out with maximizing sort of the return on the engineering dollars that you're going to put into it. And that's where we start off here. Uh We have, you know, the global flash management, the ability to manage the flash across the entire system.
14:22
We call this, I call it the Pyramid Evergreen Simple, the pyramid. There's a great blog which talks about how do we manage data inside of our systems? How do we manage read data and how do we manage right, data inside of our systems. And that really goes back to this idea of unstructured data, unstructured workloads. It's a great little blog series evergreen,
14:40
which we will talk about a little bit later, talks about sort of the that co comes out of the fact that we're going to be looking at this equipment and looking at the value you're going to get out of it over a long period of time. And if you've already used pure, you know, we focus ourselves on simplicity and we're going to carry that model through as well.
14:57
But at the end of the day, you still have storage underneath all of this. And that's what we come up with our direct flash modules. And the real focus around those direct flash modules goes right into this core of us being able to manage the individual NAN chips themselves. So deep diving a little bit more onto those direct flash modules,
15:18
uh There was a, a session um I think yesterday called Better Science. Did anybody go see that? Ok. Well, I think maybe we, we have a video recording available later on. Uh but it talks quite a bit around, around the underlines of flash uh and NAN chips. Um, if you think about a A an SSD, right? Some people are looking at saying,
15:36
oh, well, I, I already have an array. It works great with hard drives. All I gotta do is swap those out with S sds. Well, that will help you a little bit. But if you look inside of an SSD, if you pop one open, what you're gonna see in there is a couple of things, you're gonna see a lot of extra chips inside of it because they need to manage failures for
15:53
chips and failures for parts of chips, you're gonna see memory thrown inside of those because in the operations of, of that don't go nice and smooth and fast. But some of those operations like right operations take a while, they're gonna need extra pieces in there. And if you take that and you multiply it out by 50 or,
16:08
or 100 or 500 of these S sds in a large array, you end up with a system that is not nearly as efficient as you would want it to be. And uh if you know, if you're a vendor, you're just selling it. But if you're a customer, you're basically over buying a whole bunch of stuff that you don't need. And that's where direct flash comes into play.
16:26
We actually are able to address every single chip that's in every single device. And we address those individually and we address every single block within those in terms of writing data and reading data. And that gives us a little bit of an advantage because we can look holistically again, holistically across the entire environment.
16:41
Um It gives us advantages in terms of the way we can structure out the D the DF MS, it gives us advantages in terms of the way we can manage them, we can do better wear, leveling, we can do better garbage collection and all of these things. While you may say, yeah, I don't really care. I just want a storage rate. Let me just put my file some place that's fine.
16:59
But what these things do behind the scenes underneath the scenes is what they do is they give you a longer life span for the equipment, they give you more reliability for the equipment. They give you more predictability of the storage for the equipment and we're able to cut the power and the consumption of resources for you within that equipment.
17:15
And that's really where our focus was in terms of making. We said holistically, yes. If you take that SSD out and or you take that hard drive out, you plug that SSD in that price point is going to take three years to get. But if you redesign the whole system, that price point is here today and that's kind of where we focus on that.
17:33
Let's talk a little bit more about um some of the efficiencies in terms of being able to build out these DF MS, we're able to also build them out at scale better and faster than what you can get out of the hard drive markets. If you look at it today, we sell a 48 terabyte DFM in our system. I think cos probably talked about it during the keynote speech,
17:52
right? We now just announced the 75. That's a 50% boost in capacity. If you compare that roughly to three months ago, Seagate made an announcement that they just announced a 22 terabyte hard drive. You think they're going out to 50% more. A 30 terabyte hard drive really quickly.
18:11
Probably not. So we're able to get that scale up. That that's a lot better and that scale up now starts to match the very first slide that he put up there, which was that 10 X growth, not, not every customer has 10 X growth, but if you look at where you do get 10 X growth, you get 10 X growth either by collecting more data or generating more data.
18:31
Some companies collect data, all companies generate data and they all generate it in some unstructured manner because every single application out there wants to generate more data, every single person wants to debug something better and give themselves more data. And so that's where that 10 growth is starting to happen underneath the covers is the places where we don't think that it's gonna happen, but it is because somebody came out with a new
18:52
application, they want to go log 20 times more amount of data because, hey, that's, that's useful for them to debug something. Uh And again, you don't know how those workloads are gonna be. So remember, unstructured data, unstructured workloads, you need that variability in there. So I I mentioned earlier the environment and uh environmentals and you probably heard this also
19:12
before here the past few days is around, you know, power consumption and things like that. Um One of things to note is that it doesn't, it doesn't take too long to take a look out into the world, right? You got, you got wars that cause energy spikes. Uh If you went back a couple of years ago, you end up with energy spikes just due to market fluctuations.
19:28
Um And I think the the famous price, the famous comment about uh the price of gasoline, the price of energy, right? It goes up like a rocket, it comes down like a feather. So uh you know, nobody's ever in a rush to cut prices, to, you know, to, to commodities.
19:42
Um And so that's one of the things that we're seeing now where it really affects customers is this this idea that um you know, energy consumption matters and it's not so much the energy consumption of the machine itself right. That's easy to measure. And that's an easy place for us to talk about it. And uh you know, we posted our own results on the Energy Star website as well.
20:01
So we've gone through Energy Star Star Certification, but it's all the other adjunct areas that come back and affect you. Right? More energy consumption or, or more energy consumption on the array means more cooling you need in the data center. It means you needed to buy a larger, a larger air conditioning system to chill or chillers.
20:21
It means you need to get larger feeds from the power company and they charge you more for that. So all of those little things kind of add up to where? Oh yeah, this thing might use 100 extra watts. We don't care. Well, those extra 100 watts might end up costing your c your company thousands of dollars over the life span of that. And that's where we try to focus that message
20:39
and we try to make that message a little bit broader. The other place where that comes into play is you yourself, right? You have projects, you want to do things. Um I know when I worked in it, we always had projects and it was always painful when you had this nice project. It was all lined up.
20:53
You wrote up all your justification and you took it up to the boss and the boss says, yeah, that's great. We're gonna do this. And a week later you're like, hey, boss, did you sign that po, and they're like, no, we had, we had to put your project on hold, you know, maybe we'll, we'll come back and visit again in two months or three months or
21:07
four months. And that's quite painful. Right. You, you put a lot of work into that type of stuff. And so places where we're starting to see it is not directly in cancellations, it's delays, it's, it's we're going to push things out, it's going, we're slow roll things and that's where those impacts start to come into play.
21:27
Um Kind of focused on this a little bit more, you know, within flash blade, we have two platforms out there. The flash blade S and we have the flash blade E the flash blade S is focused more on those performance workloads. We're really focused on that scale up performance. Every time you add a blade, you're adding more
21:42
performance. It's great for parallel high concurrency workloads. It's great for environments where you don't know quite what the workload is going to be, where it's going to be highly variable. Uh It's a very flexible system, it's also very energy efficient. And we really targeted when we came out with this project last year and focusing on energy
21:59
efficiencies, what we then did was we kind of doubled down on that same message. And when we came out with the flash blade E earlier this year we doubled down on that again and we got there, we got there as well as targeting the hard drive market by saying, hey, look, we're going to look at this at a holistic system level and we're going to do it on a large scale. And then that way we can get the most
22:19
efficiencies out of the system, both economically environmentally and energy wise as well. And so within the flash blade environment, we have two products, the flash blade s really built for those scale up type of performance. And the flash blade also also built for scaling out but much more around workloads around more of a repository.
22:39
Both of which are very focused on sustainability, both of which we have a lot of data to back up the sustainability claims in there. And um both of which we also include within the pier one environment in terms of sustainability. So when we start talking about uh you know, was talking about the unstructured data and uh the environment we're gonna try to build that in.
22:59
We wanted to talk a little bit about sort of how do we future proof a platform and the flash blade system itself, we, we designed the chassis to be long life. So the chassis, uh if you looked at the first generation of flash blade, we designed that for an approximately 8 to 10 year life span and we're going and we'll be going on 11 years by the time it actually goes end of life.
23:20
So it's a nice life long, uh you know, long span uh chassis. The other thing we did when we went to the flash blade was we, we did the same approach to the blades. So if you're familiar with the flash array system where you can do, you know, controller upgrades every three years, we're looking at doing the same exact thing for
23:35
the blades. And so the chassis were designed to go across multiple generations of blades. The same with the networking, we already have planned bumps and speeds for networking and for the blades already within that same chassis over the lifespan of multiple blades. So we really have designed this product not for today, but what's it going to be three
23:55
generations from now is really where we were looking at it. That's also true for us to be able to put the DF MS and the higher capacity in them as well. And all of that still stays underneath this bundle of easy to manage, simple to use. You don't need to learn anything new or anything complicated about it, but you want to talk about some other aspects.
24:14
You got uh one of the things which is really important about Evergreen is the aspect of the experience that it delivers for the users. Now, I'll take you back to the keynote yesterday morning. And one of the sentiments that I took away was as a company, we take our responsibility of delivering a world class experience to our customers very,
24:39
very seriously. And our Evergreen subscription services play a huge part of that. And this actually is a unique differentiation for pure. And he touched a little bit on some of the elements that makes the, uh you know, allows us to develop uh this subscription service in a unique differentiated
25:03
manner. You can ask me like, you know, hey, why can't like, you know, everyone else do exactly the same, like, you know, evergreen subscription service. It's, it goes back to the fundamental way about how we build our product. It goes back to the DFM technologies that Andy was touching on how the hardware and software
25:21
is designed for the ground up to deliver a superior experience. What is that experience? Let's start off with the hardware portion in there. You use your storage arrays, your requirement grows over time or like, you know, you require higher capacity, higher performance or it just is a matter of using
25:43
your system where you, there is a little bit of debt debt built into the system. You just want to get in on a better system. You have the opportunity of going in and upgrading either periodically or on demand to the next version of the arrays. And that's true for the new higher denser DF MS when they are available as well.
26:02
And the second element of this is the idea of like, you know, going back on the application sprawl, there are new features, functionalities that are required and as we bring them into the system, you have the ability of upgrading your software and getting those capabilities at no additional cost non disruptively into the system as well.
26:27
And I'll go back on the non disruptive element of it because that is very, very important so that your system remains forever modern in a predictable way, your support costs are flashed and fair and you have a tremendous experience with the product, a product which remains always modern and a product which is always better tomorrow than what you buy today.
26:56
And that's what makes pure and pure product differentiated. And flash blade is no different than that. To talk a little bit about the specific elements of evergreen. Let me hand it back to Andy. So, so earlier, when I, when I mentioned the idea of the unstructured uh data equals
27:16
unstructured workload. Um And you know, the whole title of this was file, why, why put the file an object on this platform in a sense? Um If you kind of look at those types of workloads, they're not necessarily the most predictable. Sometimes you can kind of predict,
27:29
oh, this is gonna grow really fast, but sometimes you don't know when you're gonna get it or you may get it really quickly, you may have to scale up something very quickly. And so you know how, how many people in the room here are the guys, the people who are involved heavily in uh purchasing, right? And, and how many of you have some kind of budget that you have to work with?
27:48
Right? Almost everyone, right? Ok. So what we try to do here is we try to make it flexible in kind of two different dimensions. One was of course, what, what's your vehicle for doing for, for purchasing the system? But also what does that flexibility give you? So if you look at what this started off with, right?
28:04
You talked about sort of an asset right down there, which is a, a simple Capex order, right? Something costs X amount of dollars, you pay X amount of dollars you get it. Um And that's where, where the asset begins. But then what b also talked about was the plus subscription. And that's where we get into the evergreen life
28:19
cycle, which is this idea of a subscription to storage. When you, when you get a storage array, you're also buying a subscription. So that every few years, we're automatically upgrading the components inside of it. So you always have a modern storage array to hold your data that really flows very well because you know the life span of your data never aligns exactly to the life span of your
28:39
array. Most of the time the data is going to outlive the array, you're going to have to get new equipment in there. Why go through all that pain. Why not just order it as a subscription and we can refresh that hardware for you. Uh But that still focuses on those Capex type of things,
28:53
right? You have a budget of X dollars, you're gonna spend X dollars. The other thing we've done though is we've given people some flexibility with this idea of uh a subscription system that we call Evergreen flex and in that environment, that's more of saying, hey, look, you know, you want to get the hardware on your site,
29:07
but you're not going to use everything day one, you're gonna grow into it. So we're gonna give you a flexible operation in there around terms of the procurement of the hardware versus the procurement of licensing in order to use the the capacity on it, that gives people a very good flexible way of growing into a project where maybe they kind of know what the state is gonna look like, but they're not going to get their day one.
29:29
And then we went one step further, which was people are, are used to, you know, in it. You're all starting to hear about, hey, well, we could just go off to the cloud and go get something, we just pay for the consumption out on the cloud. So we wanted to do the same exact thing and put that on premise for you.
29:42
And in which case, we call that Evergreen one, that's again a consumption very similar to a cloud based consumption where as you use it, you can then sign on to, to contracts for a certain period of time at certain pricing points, you have buffer space automatically built into the systems. We give you the systems with extra buffer space. Um And it's a very flexible system in terms of
30:00
how the pricing goes. Um And these systems both come with, with buffer space built into them. So that's a really nice thing in terms of how you're going to do that. The end result here is we're giving you both flexibility on the consumption of it from a, from a spending sense, but also flexibility in terms of how do you get the equipment into your
30:19
environment to be able to handle those projects that grow faster than, than you expect or those projects that are equally unstructured in terms of their unpredictability. Thanks Andy, one of the things I just mentioned when we were talking about Evergreen is this whole idea around the experience.
30:38
And the thing that how we measure that is really our net promoter score. Uh If you, I know many of the people in here are pure customers. So you are familiar about our net promoter score industry's leading uh net promoter score, which is validated by an independent third party agency. So it's not something that, you know, we govern ourselves.
31:01
We actually make sure that, you know, it's independently verified, but this is just not a number or just for marketing purposes, we actually are truly held responsible in every element of our business to make sure that we are driving a great experience for our customer and everything we just mentioned about Evergreen is a big driver of this, but it's also a big reflection of us
31:28
intending to keep the score as high as possible, not changing tracks. One of the areas I mentioned early on is like, you know, what was the, you know, the breaking out of this traditional rigid buying life cycle. Now, if any of you like, you know, have been in storage for a decent amount of time, this should sound very familiar. On year zero,
31:57
you go in and acquire a product, you know, you start, you know, your amortization of it, operate it for about three years. Hopefully everything goes well. You like your product if you don't like it, tough luck, it's just you're stuck with it for three years or so around year three or four, you start figuring out what the future plan is for this array.
32:20
Am I going to go in and make a choice of replacing this with a new array of the same vendor or make a switch and like, you know, somewhere in around year four or five, you make the plunge plan a whole detail. Few days, you know, and like, you know, weeks of your life becomes miserable planning for it, go through the pain of going through a disruptive migration and you have a new array
32:46
and the clock resets and you are back to square one and you go through that over and over again. And that simply was like, you know, I was in a conversation with a customer in another session and while preparing and we were talking about Evergreen, I casually asked him, like, as we were talking about it, like, what do you really like about Evergreen? He's like, you know, what, every three years, I don't have to think about what I'm refreshing.
33:11
Like, that's as simple as that. Like, I get it. You, the system would work. But I just, it's just one less thing that I don't have to worry about. And the second thing he said was like, you know, the first time I had to do a migration, I sent an email to every single person in my company.
33:27
We prepared for it over a weekend. Everybody was supposed to expect a like, you know, uh some sort of outage, like, you know, unavailability of service. We did that over the weekend. The second time I got gutsy. It was like a weekday evening, the third time. It was like,
33:44
you know, on a Wednesday at 9 a.m. Let's do it. I don't really care. It would be, it would be good. But that's the kind of a difference between this life cycle and the one that we believe is the right approach in the current age.
34:00
You have a product for in our flash entire pure stack, including our flash blade family. Whether you are looking at flash blade S and D, you can acquire it as a Capex or as a storage as a service. You in a you know, in a flexible manner using evergreen flex or evergreen one, your business requirements change, you can upgrade it on demand or like as part of evergreen forever.
34:24
Periodically you can upgrade that you require more capacity. You have the ability of just increasing your capacity, your performance requirements change, you can upgrade your blades, you need any kind of new features and functionalities that we are developing. They are available in all inclusive as part of your software stack.
34:46
And guess what? We talked a little bit about the efficiency and the density of the DFM you want to go in and become more power efficient. You have the ability of doing that as well. The same customer that I was talking about when we asked him the question around like, you know, the importance of power and space efficiency.
35:07
His response was the best answer I can give you now is I started with one array, one use case I support about seven or eight use cases right now. And my footprint isn't any different than what it looked like when I first deployed the pure, pure my data has grown. So I mean, yes, there are other tangential results.
35:35
But if you want to visualize it. My data has grown many folds. My use cases have grown many fold. The number of products I have has has doubled or tripled, but it still is the same experience. And that's the idea of disrupting the legacy rigid life cycle of buying storage that we believe is the right way of going forward.
36:01
Now to summarize, I asked five areas of like, you know what we hear as the challenge for the customers. What does Flash Blade bring to the picture? You have a scenario where storage really enables your data and your application to support whichever business decisions that you need to make and your organization needs to make. You have an architecture which is incredibly
36:31
simple to manage and can handle growth in capacity and performance seamlessly and non disruptively. Thirdly, you have the industry's leading and most power efficient storage solution across the bone. Fourthly, you have a platform which will protect your storage investments today and over time as your requirements change.
37:04
And finally, you have this system which breaks the buying cycle so that you never have to go through a down time ever again. So that's all I had. Happy to stay back here for some questions and I wanted to thank you again for, you know, spending time with us so late during accelerate but happy to take questions in here
37:28
or like, you know, if you have more detailed conversations, both Andy and I will be in the uh zone, uh after this. So, you know, we can continue the conversation there as well.
