 Live from San Diego, California, it's theCUBE. Covering KubeCon and CloudNativeCon, brought to you by Red Hat, the CloudNative Computing Foundation and its ecosystem partners. Welcome back to theCUBE here in San Diego for KubeCon and CloudNativeCon 2019. It's our fourth year of doing theCUBE here. I'm Stu Miniman, it's my fourth time I've done the show. Joining me is Justin Warren. He's actually been to more of the KubeCon than the Kube has, I think, at least in North America. And welcome into the program. Two veterans of these events from Pure Storage, sitting to my right is Shilpi Shrivastava, who's a director of product marketing and sitting to her right is Anthony Laferrario, who's a senior product manager, both of you with Pure Storage. Thank you so much for joining us. Thanks for having us. Thanks for having us, yeah. All right, so we're kind of joking about veterans here because we know that things are moving faster and faster. You both work for Storage Company. Storage is not known to be the fastest moving industry. It's been fascinating for me to watch kind of things picking up, the pace of change, especially when you talk about how developers and software in a multi-cloud environment fit out. So, Shilpi, maybe give us a frame for, you know, you're in a Kubernetes t-shirt here, Pure's at the show. How should we be thinking about Pure in this ecosystem? Sure, yeah, so Pure, as you know, we started off as an all-flash, on-prem Storage Company 10 years ago. And we've kept pace with constantly innovating and making sure we're meeting our customers' needs. One of the areas, of course, that we see a lot of enterprises moving today is towards microservices, towards containerized applications. And our goal at Pure really is to help customers modernize their applications while still keeping that storage seamless and keeping that invisible to the application developers. I think it actually lines up really well. If you look just at Pure's sort of steam across time has been performance with simplicity, right? And I think the simplicity argument starts to mean something different over time, but it's a place that we still want to really focus. As our customers started to use, try to containerize their applications, there are a couple of challenges we saw. Now, container environments, of course, they're known for their agility, how portable they are, their lightweight, and their fast. And when they're fast, storage can sometimes be a bottleneck because your storage might not necessarily scale as fast. It might not be able to provision storage volumes as fast to your container environments. And that's the challenge that we at Pure wanted to solve with our Kubernetes integrations. Anthony, you mentioned simplicity there, so I'm going to challenge you a bit on that because Kubernetes is generally not perceived as being particularly simple. And the storage interfaces as well, like stateful sets is kind of only really stabilized over the last, say, 18 months. So how is Pure actually helping to make the Kubernetes experience simpler for developers? Yeah, and you know, you're totally right. I don't think I was necessarily saying that someone looking for the simplest thing they could ever find would adopt Kubernetes and expect to find that. But what I really meant was, on one hand you have your more traditional enterprise infrastructure type folks who are trying to build out the underlying private cloud that you're going to deploy your infrastructure on. And on the other hand, you have your developers, you have your Kubernetes, you have your cloud native applications, and really the interface between those is where I'm looking at that simplicity argument because traditionally Pure has focused on that simple interface to the end user. But the end user, as we were talking about before the show, has shifted from a person to being a machine. And the objective for Pure and what we're building on the cloud native side is how do we take that simple sort of as-a-service consumption experience and present that on top of what looks like a traditional infrastructure platform. So I can get more into the details of that if you'd like, but really that layer is where we're focused on the simplicity and really just asking the end user as few questions as we can. I just want to ask you, what do you need? I don't want to ask you, tell me about the IQN and blah, blah, blah. They don't want that, right? That's the simplicity I'm talking about. Yeah, you're right. Developers generally, the idea of DevOps and I challenge people whenever they mention DevOps and I'm hearing a pretty consistent message that developers really don't care about infrastructure and don't want to have anything to do with it at all. So if you can just bake it into the system and somehow make it easier to operate at that kind of SRE level, that infrastructure level, that Kubernetes as a platform. So once that's solved, then as a developer, I can just get on with writing some code. Developers definitely want storage to be invisible. Yeah. So if you want, but if they want storage to be invisible, that's not so great for your brand because you actually want them to know and care about having a particular storage platform. So how do you balance that idea that we want to show you that we can have innovative products that you care about the storage, but you also don't need to care about the storage at all because we'll make it invisible? How does that work? So storage for container environments has been a challenge and what we are trying to educate the platform level users is that with the right kind of storage, it can actually be easy. Storage for Kubernetes can be easy and the way we make it simple or invisible is through the automation that we provide. So Pure Service Orchestrator is our automation for storage delivery into the containerized environments and so it's delivered to a CSI plugin, but we try to do a little more than just develop a plugin into your Kubernetes environments. We try to make your scalability seamless. So it's super easy to add new storage and so yeah, I think because container environments were initially developed for states less applications, when we came to stateful applications, there's still a thing about, oh, why should I care about storage? But people are slowly realizing that we need to care about it because we don't want to ultimately be bothered by it, right? And if I can make a point to just tag on to that, the conversations I've had at the show this week have even helped me sort of crystallize the way I like to explain this to people, which is at first, a lot of people will say, oh, I don't do stateful applications, I'm doing state less applications in Kubernetes and my response is okay, I understand that you've decided to externalize the state of your system from your Kubernetes deployment, but at some point you have to deal with state. Now whether that's an oracle database you happen to be calling out to outside of your Kubernetes cluster, whether that's a service from a public cloud like S3 or whether that's deciding to internalize that state into Kubernetes and manage it through the same management plane, you have to have state. Now, when we talk about what we're doing in PSO and why that's valuable and why, to your point about the brand, I don't necessarily worry, is because when we can give a seamless experience at the developer layer and we can give the SRE or the cluster manager layer a way where they can have a trusted high performance, high availability storage platform that their developers consume without knowing or worrying about it and then as we look into the future, how do we handle cross cluster and multi-cloud stateful workflows? We can really add value there, so. Yeah, I'm glad you brought up the multi-cloud piece of it because one of the more interesting things I saw from Pure this year is how Pure is putting its software cloud native. So when I saw that, one of the questions like, okay, when I come to a show like this, how does Kubernetes and containers fit into that old discussion? So help us connect the dots as to what was announced and everything else that's happening. Yeah, so you've heard about Cloud Blockstore, which is our software running on the AWS Cloud today and that's basically what we've done is we've, people have loved FlashArray all these years for the simplicity it provides for the automation and performance. We wanted to give you something similar and something enterprise great in the public cloud. So Cloud Blockstore is basically, you can think of it as a virtual FlashArray on the AWS Cloud. So with that, you now have deduplication, thin provisioning capabilities in the cloud. You can, we've brought an active cluster which is active-active synchronous replication between availability zones. So really making your AWS environments ready for mission critical applications. Plus with our pure PSO just works the same way on-prem as in the cloud. So it's just great for hybrid application mobility. You have the same APIs on both sides. Yeah, it's actually very cool, right? One of the fun things for me as a software developer at Pure, at a software side guy at Pure is that the APIs that our arrays have are the same APIs. It's actually the same underlying software version even though it's a totally different hardware backend implementation when we run in a cloud-native form factor versus when we run in a physical appliance form factor. The replication engines work between the two. Snapshots, clones, our ability to do instant restores. Like everything that we do and that has brought value from our storage software stack, we still get access to in a cloud-native environment and the transports as well, right? I guess trying to understand is there Kubernetes involved here or is this just natively in AWS and then on-premises and you've got APIs in between? So Kubernetes itself is a compute orchestration layer component. So when I look at Kubernetes, I would say Kubernetes sits above both sides, right? Or potentially above and across both sides depending on how you decide to structure your environment. But the nice part is if you develop a cloud-native application, right? And that's running on Kubernetes. The ability to support that with the same storage interface is the same SLAs, move it efficiently, copy it efficiently and do that on whatever cloud you care to do. That's where it gets really cool. So we developed this really cool demo where you have a container application running on PSO, on Flash, or on-prem. We migrated that to cloud block store on AWS and it just runs. You use the same YAML scripts in both places. There's no need to do a massive re-architecture or anything. Your application just runs when you move it and we take care of all the data mobility with our asynchronous replications. You can take a snapshot on-prem, you can snap it out into AWS, restore it back into cloud block store. So it really opens up a lot of new use cases and makes them simple for customers. That idea of write once, run anywhere. I'm old enough to remember when Java was a brand new thing and that was the promise and it never quite got there because it turns out it's really, really hard to do that. But we are seeing from Pure and from a lot of vendors here at the show that there's a lot of work and effort being put in into that difficult problem so that other people don't have to care about it. So you're building that abstraction in and working on how the details of this work. And I was fortunate enough to get a deep dive into the architecture of cloud block store at recent Accelerate conference and the way you've actually used cloud resources as if they were kind of infrastructure components and then built the abstraction on top of it. But in the same way that it runs on-site, that's what gives you that ability to keep everything the same and make it simple, doing a lot of hard work and hard engineering underneath so that no one has to care anymore. Yeah, and the way we've architected cloud block store is that we've used the highest performing AWS infrastructure and the highest durability AWS infrastructure. So you're actually now able to buy performance and durability through one single virtual appliance. Yeah, how's the adoption of the products going? I know it was very early when it was announced just a few months ago. So what's the feedback from customers been so far about global cloud block store? It's been really positive. And actually, the one use case that I want to highlight really most is actually DevOps use cases, right? The value add of being able to have the same deployment for that application, for a test or dev infrastructure in one cloud versus a production deployment in another cloud has been very exciting for folks. So when you think about that use case in particular, the ability to say, okay, I'm coming up to a major quarterly release or whatever I have for my product, I need to establish a bunch more test environments. I don't necessarily want to have bought that. And we're not necessarily talking about bursting over the wire anymore, right? We're talking about local storage under the same interfaces in the cloud that you choose to spin up all of those test environments. So cases like that are pretty interesting for folks. Yeah, I think that's how people have started to realize that it's that operation side of things. It's not even day two, it's day 90 and day 147 where I want to be operating this in the same place, in the same way, no matter where it is, because it just saves me so much heartache and time of not having to re-implement different interfaces. And I don't have to retrain my resources because it all looks the same. So, yeah, DevTest definitely a big use case. Migration through VWalls, that's another use case that we're seeing a lot of customers interested in. And disaster recovery, using AWS as a disaster recovery. How do you, so you can efficiently store backups on Amazon S3, but how do you do an easy, fast restore to actually run your applications there? So with Cloud Block Store, it is now possible to do that, to do a fast, easy restore. Also, a couple of weeks ago actually, we started taking registrations for our beta program for Cloud Block Store for Azure as well. Yep, customers are going multi-cloud, we are going multi-cloud with them. Great, want to give you both final word takeaways for pure storage participation here at the show. I think the biggest thing that I want people to understand, and I actually gave this talk at the Cloud Native Storage Day on day zero, is that Cloud Native Storage is an approach to storage. It's not a location for storage. And I think pure storage, that really defines to me the way we're going about this. We're trying to be Cloud Native Storage, wherever you need it. So that's really the takeaway I'd like people to have about pure. And storage for Kubernetes doesn't have to be hard. We are here all day today as well. So I mean, this is a challenge the industry is seeing today and we have a solution to solve that for you. All right, well, that's a bold statement to help end this. Shilpi, Anthony, thank you so much for joining us. Thank you guys. Thank you so much for having us. For Justin Warren, I'm Stu Miniman, back with more coverage here from KubeCon Cloud Native Con 2019. Stay classy San Diego and thanks for watching the Kube.