Prakash Darji, Pure Storage | At Your Storage Service
(bright intro music) >> The cloud has popularized many useful concepts in the past decade, working backwards from the customer to pizza teams and DevOps mindset, the shared responsibility model, and security, of course, the shift from CapEx to OPEX, and as a service consumption models. The last item is what we're here to talk about today. Pay for consumption is attractive because you're not over provisioning, at least not the way you used to. You'd have to buy for peak capacity events, but there are always two sides to every story, and while pay for use more closely ties IT consumption to business value, procurement teams don't always love the uncertainty of the cloud bill each month, but consumption pricing and as a service models are here to stay in software and hardware. Hello, I'm Dave Vellante and welcome to "At Your Storage Service" made possible by Pure Storage, and with me is Prakash Darji who's the General Manager of the Digital Experience Business Unit at Pure. Prakash, welcome to the program. >> Thanks Dave. Thanks for having me. >> You bet. Okay, we've seen this shift to as a service, the as a service economy, subscription models, and this as a service movement have gained real momentum. It's clear over the past several years. What's driving this shift? Is it pressure from investors and technology companies that are chasing the all important ARR, their annual recurring revenue stream? Is it customer driven? Give us your insights. >> Well, look, I think we'll do some definitional stuff first. I think we often mix the definition of a subscription and a service, but, you know, subscription is, "Hey, I can go for a pay up front or pay as I go." Service is more about, "How do I not buy something just by the outcome?" So, you know, the concept of delivering storage as a service means, what do you want in storage, performance, capacity, availability? Like that's what you want. Well, how do you get that without having to worry about the labor of planning, capacity management? Those labor elements are what's driving it. So I think in the world where you have to do more with less and in a world where security becomes increasingly important where standardization will allow you to secure your landscape against ransomware and those types of things, those trends are driving the SaaSification of storage, and the only way to deliver that is storage as a service. >> So that's good. You maybe thinking about it differently than some of the other companies that I talked to, but so you've made inroads here, pretty big inroads actually, and changed the thinking in enterprise data storage with a huge emphasis on simplicity. That's really Pure's raison d'etre. How does storage as a service fit in to your innovation agenda overall? >> Well, our innovation agenda started, as you mentioned, with the simplicity, you know, a decade ago with the Evergreen Architecture. That architecture was beyond the box. How do you go ahead and say, "I can improve performance or capacity as I need it." Well, that's a foundational element to deliver a service because once you have that technology, you can say, "Oh, you know what? You've subscribed to this performance level. You want to raise your performance level and yes, that'll be a higher dollar per gig or dollar per terabyte," but how do you do that without a data migration? How do you do that with a non-disruptive service change? How do you do that with a delivery via software update? Those elements of non-disruptive updates, when you think SaaS, Salesforce, you don't know when Salesforce doesn't update. You don't know when they're increasing something, adding a new capability. It just shows up. It's not a disruptive event. So to drive that standardization and SaaSification in service delivery, you need to keep that simplicity of delivery first and foremost, and you can't allow, like, if the goal was, "I want to change from this service here to that service here," and a person needed to show up and do a day data migration, that's kind of useless. You've broken the experience of flexibility for a customer. >> Okay, so I like the Salesforce analogy, but I want to jump out do a little side for a second. So I've got to make some commitment to Pure, right? Some baseline commitment, and if I do, then I can dial up and then pay for what I use, and I can dial it down, correct? >> Correct. >> Okay. I can't do that with Salesforce, all right? I could dial up, but then I'm stuck with those licenses. So you have a better model in Salesforce, I would argue. Okay. >> Yeah. I would agree with that. >> Okay, so, and I got to pay for everything up front. Anyway, let's go back to I was kind of pushing at you a little bit at my upfront, you know, about, you know, the ARR model, the all-important, you know, financial metric, but let's talk from the customer standpoint. What are the benefits of consuming storage as a service from your customer's perspective? >> Well, one is when you start your storage journey, do you really know what you need? And I would argue, most of the time, people are guessing, right? It's like, "Well, I think I need this. This is the performance I think I need or this is the capacity I think I need," and, you know, with the scientific method, you actually deploy something, and you're like, "Do I need more? Do I need less?" You find out as you're deploying. So in a storage as a service world, when you have the ability to move up performance levels or move out capacity levels, and you have that flexibility, then you have the ability to just to meet demand as you deploy, and that's the most important element of meeting business needs today. The applications you deploy are not in your control when you're providing storage to your end consumers. >> Yeah. >> They're going to want different levels of storage. They're going to want different performance thresholds. That's kind of a pay, you know, pay for performance type culture, right? You can use HR analogy for it. You pay for performance. You want top talent, you pay for it. You want top storage performance, you pay for it. You don't, you can pay less, and you can actually get lower performance tiers. Not everything is a tier one application, and you need the ability to deploy it, but when you start, how do you know the way your end customers are going to be consuming or do you need a dictated upfront? 'Cause that's infrastructure dictating business inflexibility, and you never want to be in that position. >> I got another analogy for you. It's like, you know, we do a lot of hosting at our home and you know, like Thanksgiving, right? And you go to the liquor store and say, "Okay, what should I get, should we get red wine? We got to go white wine, we got to get some beer. Should I get bubbles? Yeah, I get some bubbles." 'Cause you don't know what people are going to have, and so you over provision everything and then there's a run on bubbles and you're like, "Ah, we're running the bubbles," so you just over buy, but there's a liquor store that actually will take it back. So I got to do business with those guys every time 'cause it's way more flexible. I can dial up capacity or I can dial up performance, and dial it back down if I don't use it. >> Where you're going to be drinking a lot more the next few weeks. >> Yeah, exactly, like which is the last thing you want. Okay, so let's talk about how Pure kind of meets this as a service demand. You've touched upon your differentiators from others in the market. You know, love to hear about the momentum. What are you seeing out there? >> Yeah, look, our business is growing well largely built on, you know, what customers need. Specifically, where the market is at today is there's a set of folks that are interested in the financial transformation of CapEx to OPEX. Like that definitely exists in the industry around, "How do I get a paper use model?" The next kind of more advanced customer is interested in, "How do I go ahead and remove labor to deliver storage?" And a service gets you there on top of a subscription. The most sophisticated customer says, "How do I separate storage production with consumption and production of storage?" Being a storage producer should be about standardization so I could do policy based management. Why is that important? You know, coming back to some of the things I said earlier, in the world where ransomware attacks are common, you need the standardized security policies. Linux has new vulnerabilities every other day, like find two, three critical vulnerabilities a week. How do you stay on top of it? The complexity of staying on top of it should be, "Look, let's standardize and make it a vendor problem, and assume the vendor's going to deliver this to me." So that standardization allows you to have business policies that allow you to stay current and modern. I would argue in, you know, the traditional storage and appliance world, you buy something and the day after you buy it, it's worthless. It's like driving a car off a lot, right? The very next day, the car's not worth what it was when you bought it. Storage is the same way. So how do you ensure that your storage stays current? How do you ensure that it gets a like a fine line that gets better with age? Well, if you're not buying storage and you're buying a performance SLA, it's up to the vendor to meet that SLA so it actually never gets worse over time. This is the way you modernize technology and avoid technology debt as a customer. >> Yeah, I mean, just even though words you're using and the way you're thinking about this precaution, I think are different, and I love the concept of essentially taking my labor cost and transferring them to Pure's R&D. I mean, that's essentially what you're talking about here. So let's stick with the tech for a minute. What do you see as new or emerging technologies that are helping accelerate this shift toward the as a service economy? >> Well, the first thing is I always tell people you can't deliver a service without monitoring because if you can't monitor something, how you're going to know whether you're meeting your service level obligation, right? So everything starts with data monitoring. The next step layering on the technology differentiation is if you need to deliver a service level obligation on top of that data monitoring, you need the ability to flexibly meet whatever performance obligations you have in a tight time window. So supply chain and being able to deliver anywhere becomes important. So if you use the analogy today of how Tesla works or a IoT system works, you have a SaaS management that actually provides instructions that pushes those instructions and policies to the edge. In Tesla's case, that happens to be the car. It'll push software updates to the car. It'll push new map updates to the car, but the car is running independently. It's not like if the car becomes disconnected from the internet, it's going to crash and drive you off the road. In the same way, what if you think about storage as something that needs to be wherever your application is? So people think about cloud as a destination. I think that's a fallacy. You have to think about the world in the view of an application. An application needs data, and that data needs to sit in storage wherever that application sits. So for us, the storage system is just an edge device. It can be sitting in your data center it can be sitting in a Equinix. It can be sitting in hosted and MSP can run it. It can even be sitting in the public cloud, but how do you have central monitoring and central management where you can push policies to update all those devices, very similar to an IOT system? So the technology advantage of doing that means that you can operate anywhere and ensure you have a consistent set of policies, a consistent set of protection, a consistent set of, you know, prevention against ransomware attack regardless of your application, regardless of, you know, where it sits, regardless of what content in it you're on. That approach is very similar to the way the IoT industry has been updating and monitoring edge devices, nest thermostats, you know, Tesla cars, those types of things. That's the thinking that needs to come to storage, and that's the foundation on which we built Pure as a service. >> So that implies or, at least I infer, that you've, obviously, got control of the experience on prem, but you're extending that into AWS, Google, Azure, which suggests to me that you have to hide the underlying complexity of the primitives and APIs in that world, and then eventually, actually today, 'cause you're treating everything like the edge out to the edge, you know, maybe mini Pure at some point in time, but so I call that super cloud, that abstraction layer that floats above all the clouds on-prem and adds that layer of value and is a singular experience, what you're talking about pushing, you know, policy throughout. Is that the right way to think about it? And how does this impact the ability to deliver true storage as a service? >> Oh, that's absolutely the right way of thinking about it. The things that you think about from an abstraction kind of fall in three buckets. First, you need management. So how do you ensure consistent management experience, creating volumes, deleting volumes, creating buckets, creating files, creating directories like management of objects and create a consistent API across the entire landscape? The second one is monitoring. How do you measure utilization and performance obligations or capacity obligations or, you know, policy violations, wherever you're at? And then the third one is more of a business one, which is procurement because you can't do it independent of procurement, meaning what happens when you run out? Do you need to increase your reserve commits? Do you want to go on demand? How do you integrate it into company's procurement models such that you can say, "I can use what I need," and any, it's not like every change order is a request of procurement. That's going to break an as a service delivery model. So to get embedded in a customer's landscape where they don't have to worry about storage, you have to provide that consistency on management, monitoring, and procurement across the tech, and yes, this is deep technology problems, whether it's running our storage on AWS or Azure or running it on prem or, you know, at some point in the future, maybe even, you know, Pure mini at the edge, right? So, you know, all of those things are tied to our Pure as a service delivery. >> Yeah, technically, non-trivial, but hey, you guys are on it. Well, we got to leave it there, Prakash. Thank you, great stuff, really appreciate your time. >> All right, thanks for having me, man. >> You're very welcome. Okay, in a moment, Steve McDowell. For more insights and strategies, he's going to give us the analyst perspective on as a service. You're watching "theCUBE", the leader in high tech enterprise coverage. (bright outro music)
SUMMARY :
at least not the way you used to. Thanks for having me. that are chasing the all important ARR, So, you know, the concept and changed the thinking and you can't allow, So I've got to make some So you have a better model I would agree with that. the all-important, you and you have that flexibility, and you need the ability to deploy it, and you know, like Thanksgiving, right? a lot more the next few weeks. like which is the last thing you want. This is the way you modernize technology and the way you're thinking and ensure you have a out to the edge, you know, such that you can say, but hey, you guys are on it. the leader in high tech
SENTIMENT ANALYSIS :
ENTITIES
Entity | Category | Confidence |
---|---|---|
Dave Vellante | PERSON | 0.99+ |
Steve McDowell | PERSON | 0.99+ |
Dave | PERSON | 0.99+ |
two | QUANTITY | 0.99+ |
Tesla | ORGANIZATION | 0.99+ |
Pure Storage | ORGANIZATION | 0.99+ |
OPEX | ORGANIZATION | 0.99+ |
Prakash Darji | PERSON | 0.99+ |
two sides | QUANTITY | 0.99+ |
Prakash | PERSON | 0.99+ |
First | QUANTITY | 0.99+ |
ORGANIZATION | 0.99+ | |
AWS | ORGANIZATION | 0.99+ |
CapEx | ORGANIZATION | 0.99+ |
second one | QUANTITY | 0.99+ |
first thing | QUANTITY | 0.98+ |
today | DATE | 0.98+ |
third one | QUANTITY | 0.98+ |
each month | QUANTITY | 0.97+ |
Thanksgiving | EVENT | 0.97+ |
Equinix | ORGANIZATION | 0.96+ |
a decade ago | DATE | 0.95+ |
Pure mini | COMMERCIAL_ITEM | 0.94+ |
three critical vulnerabilities | QUANTITY | 0.93+ |
Salesforce | ORGANIZATION | 0.93+ |
next day | DATE | 0.91+ |
past decade | DATE | 0.91+ |
Linux | TITLE | 0.89+ |
a week | QUANTITY | 0.88+ |
first | QUANTITY | 0.85+ |
Pure. Prakash | ORGANIZATION | 0.84+ |
Azure | ORGANIZATION | 0.83+ |
Evergreen Architecture | ORGANIZATION | 0.79+ |
Azure | TITLE | 0.78+ |
Pure | COMMERCIAL_ITEM | 0.77+ |
next few weeks | DATE | 0.74+ |
years | DATE | 0.72+ |
one | QUANTITY | 0.69+ |
past | DATE | 0.65+ |
Pure | ORGANIZATION | 0.61+ |
a day | QUANTITY | 0.61+ |
SaaS | TITLE | 0.55+ |
Salesforce | TITLE | 0.51+ |
second | QUANTITY | 0.48+ |
theCUBE | TITLE | 0.45+ |