 From around the globe, it's theCUBE with coverage of KubeCon and CloudNativeCon North America 2020 virtual, brought to you by Red Hat, the CloudNative Computing Foundation and Ecosystem Partners. Hey, welcome back, everybody. Jeff Frick here with theCUBE, coming to you from our Palo Alto studios with our ongoing coverage of KubeCon, CloudNativeCon 2020 North America. Of course it's virtual like everything else is in 2020, but we're excited to be back. It's a terrific show and we're excited to have our next guest. So let's introduce him in. We've got Sam Werner, the VP of an offering manager and business line executive for storage, for IBM. Sam, great to see you. Great to be here. And also joining us is Brent Compton. He's the senior director of data services for Red Hat. Great to see you, Brent. Thank you. So let's jump into it. CloudNative, everything's about CloudNative, everything's about containers, everything's about kind of containerization and flexibility. But then there's this thing on the back and called storage. And we actually have to keep this stuff and record this stuff and have data protection for this stuff and business resiliency. Love to jump into it. So let's, you know, where does storage fit within a container world and how is the growth of containers and the adoption containers really had you rethink the way that you think about storage and how clients should think about storage. Sam, let's start with you. Yeah, I mean, it's a great question. And first off, I'm really excited about another KubeCon. We did Europe now, doing North America. So very excited to be, you know, seeing all the, you know, all the news and all the people talking about the advancements around Kubernetes. And we're very excited about it. Now, you asked a very good question, an important question. We're seeing an acceleration of digital transformation and the people that are going through this digital transformation are using containers to now modernize the rest of their infrastructure. The interesting thing about it though is those initiatives are being driven out of the application teams, the business lines in an organization. And a lot of them don't understand that there's a lot of complexity to the storage piece here. So the storage teams I talked to are all of a sudden getting these initiatives thrown on them or kind of halfway their strategy and they're scratching their heads trying to figure out now how they can support these applications with persistent storage because that's not where containers started. They started with microservices and now they're in a quandary, they have to deliver a certain SLA to their customers and they're trying to figure out how they do it in this new environment which in a lot of cases has been designed outside of their scope. So they're seeing issues with data protection. Some of the kind of core things that they've been dealing with for years are now, they're now having to solve all over again. So that's what we're working on helping them with. Reinventing how storage is deployed to help them deliver the same level of security availability and everything they have in the past in these new environments. Right. So, Brent. And Jeff. Yeah, go ahead. Well, I can say you've been involved in this for a long time. You know, you've worked in hyperconverge, you've worked in big data, you know, the evolution of big data continues to change as ultimately we want to get people the information to make good decisions but we've gone through a lot of iterations over the year. So how is it different now? How is it different with containers? What can we finally do you as an architect that we couldn't do before? Infrastructure is code. That's I think one of the fundamental differences of the storage admin of yesteryear versus the storage admin of today. Today, as Sam mentioned, as people are developing and deploying applications, those applications need to dynamically provision the infrastructure, dynamically provision what they need from compute, dynamically provision what they need from storage, dynamically provision network paths. And so that element of infrastructure as code, a dynamically provisioned infrastructure is very different from, well, from yesterday when applications or teams needed to, well, when they needed storage, they would file a ticket and typically wait. Now they make an API, now they make an API call and storage is dynamically provisioned and provided to their application. But what's I think hard to understand for the layman and maybe it's just me, right? It's very easy to understand dynamic infrastructure around compute, right? Pepsi, I'm running it out for the Super Bowl. I need, I know a whole bunch of people are going to hit my, hit my site. And it's kind of easy to understand dynamic provisioning around networking. Again, for the same example, what's less easy to understand is dynamic provisioning for storage. It's one thing to say, you know, there's a pool of storage resources that I'm going to dynamically provision for this particular app for this particular moment. But one of the whole things about the dynamic is not only is it available when you need it, but I can make it big. And conversely, I can make it smaller, go away. I get that for servers and I kind of get that for networking, supporting an application and that example I just talked about, but we can't, it doesn't go away a lot of the time for storage, right? That's important data that's maybe feeding another process. There's all kinds of rules and regulations. So when you talk about dynamic infrastructure for storage, it makes a lot of sense for grabbing some to provision for some new application, but it's harder to understand in terms of true dynamics in terms of either scaling down or scaling up or turning off when I don't particularly need that much capacity or even that application right now. How does it work within storage versus you know, just servers where I'm grabbing them and then I'm putting it back in the pool? Well, let me start on this one then I'm going to hand it off to Brent. You know, let's not forget by the way that enterprises have very significant investments in infrastructure and they're able to deliver six nines of availability on their storage and they have DR worked out and all of their security, encryption, everything it's already in place and they are sure that they can deliver on their SLAs. So they want to start with that and you have to leverage that investment. So first of all, you have to figure out how to automate that into the environment that existing SAM and that's where things like APIs, the container storage interface, CSI drivers come in IBM provides that across their entire portfolio allowing you to integrate your storage into a Kubernetes environment, into an OpenShift environment so that it can be automated but you have to go beyond that and be able to extend that environment then into other infrastructure for example, into a public cloud. So with the IBM Flash system family with our spectrum virtualized software we're actually able to deploy that storage layer not only on-prem on our award-winning arrays but we can also do it in the cloud. So we allow you to take your existing infrastructure investments and integrate that into your Kubernetes environment and using things like Ansible fully automate the environment. I'll get into data protection before we're done talking but I do want Brent to talk a bit about how container native storage comes into that mix as well and how you can start building out new environments for your applications. Yeah, what the two of you are alluding to is effectively a Kubernetes services layer which is not storage. It consumes storage from the infrastructure as Sam said, just because people deploy Kubernetes cluster doesn't mean that they go out and get an entirely new infrastructure for that. If they're deploying their Kubernetes cluster on-premises, they have servers. If they're deploying their Kubernetes cluster on AWS or on Azure on GCP they have infrastructure there. What the two of you are alluding to is that services layer which is independent of storage that can dynamically provision, provide data protection services. As Sam mentioned, we have good stuff to talk about there relative to data protection services for Kubernetes clusters but it's the abstraction layer or data services layer that sits on top of storage and which is different. The basics of storage underneath in the infrastructure remain the same, Jeff. But how that storage is provisioned and this abstraction layer of services which sits on top of the storage, storage might be IBM Flash System Array, storage may be EMC SAN, storage may be AWS EBS, that's the storage infrastructure. But this abstraction layer that sits on top of the status services layer is what allows for the dynamic interaction of applications with the underlying storage infrastructure. And then again, just for people that aren't completely tuned in, then what's the benefit to the application developer, provider, distributor, with that type of an infrastructure behind them, what can they do that they just couldn't do before? Well, I mean, look, I'll let you in there. I mean, we're trying to solve the same problem over and over again, right? It's always about helping application developers build applications more quickly, helps them be more agile. IT's always trying to keep up with the application developer and always struggles to, in fact, that's where the emergence of cloud really came from, was trying to keep up with the developer. So by giving them that automation, it gives them the ability to provision storage in real time. Of course, without having to open a ticket like Brent said, but really the holy grail here is getting to a develop once and deploy anywhere model. That's what they're trying to get to. So having an automated storage layer allows them to do that and ensure that they have access to storage and data no matter where their application gets deployed. Right, right. That pesky little detail when I have to develop that app, it does have to sit somewhere. And I don't think storage really has gotten enough of the bright light really in kind of this app-centric developer-centric world. We talk all the time about having compute available and software-defined networking, but having this software-defined storage that lives comfortably in this container world is pretty interesting and a great development. I want to shift gears a little bit. Jeff, just one thing. Yeah, go ahead. Plus one to Sam's comments there, all the application developer wants, they want an API and they want the same API to provision the storage regardless of where their app is running. Right. The rest of the details, they usually don't care about. Sure, they want it to perform what not. Give them an API and make it the same regardless of where they're running the app. I was going to say, not only do they want to perform, they probably just presume performance, right? I mean, that's the other thing is that the best in class quickly becomes presumed baseline in a very short period of time. So you've got to just deliver the goods, right? They're going to get frustrated and not be productive. But I wanted to shift gears up a little bit and talk about some of the macro trends, right? We're here towards the end of 2020. Obviously COVID had a huge impact on business in a lot of different ways. And it's really evolved from, you know, March, the slight switch moment, everybody worked from home to now this kind of extended time that's probably going to go on for a while. I'm just curious, some of the things that you've seen with your customers not so much at the beginning because that was a special and short period of time but more as we've extended and are looking to probably extend to this for a while. You know, what is the impact of this increased work from home, increased attack surface? You know, some of these macro things that we're seeing that COVID has caused and then the other kind of macro trends beyond just this containerization that you guys are seeing impacting your world. Start with you, Sam. Yeah, you know, I don't think it's actually changed what people were going to do or the strategy. What I've seen it do is accelerate things and maybe change the way they're getting there. And so there are actually a lot of enterprises are running into challenges more quickly than they thought they would. And so they're coming to us and asking us to help them, solve them. For example, backing up their data in these container environments. As you move mission critical applications that maybe we're going to move more slowly, they're realizing that as they've moved them, they can't get the level of data protection they need. And that's why actually we just announced at the end of October updates to our modern data protection portfolio. It now is containerized. It can be deployed very easily in an automated fashion but on top of that, it integrates down into the API layer, down into CSI drivers and allows you to do container aware snapshots of your applications. So you can do operational recovery. If there's some sort of an event you can recover from it, you can do DR and you can even use it for data migration. So we're helping them accelerate. So the biggest, I think requests I'm getting from our customers is how can you help us accelerate and how can you help us fix these problems that we went running into as we tried to accelerate our digital transformation? Brent, anyone that you wanna highlight? Ironically, one of my team was just speaking with one of the cruise lines two days ago. We all know what's happened to them. So if we just use them as an example, clearly our customers need to do things differently now. So plus one to Sam's statement about acceleration and I would add another word to that which is agility. Frankly, they're having to do things in ways they never envisioned 10 months ago. So their need to cut cycle times to deploy effectively new ways of how they transact business has resulted in accelerated pull for these types of infrastructure as code technologies. That's great. The one that jumped in my mind, Sam, as you were talking, we've had a lot of conversations. Obviously security always comes up and baking security in is a theme, but ransomware as a specific type of security threat. And the fact that these guys not only want to lock up your data, but they want to go in and find the backup copies and really mess you up. So it sounds like that's even more important to have this safe and we're hearing all these conversations about air gaps and dynamic air gaps and can we get air gaps in some of these infrastructure setups so that we can put those backups and recovery data sets in a safe place so that if we have a ransomware issue, getting back online is a really, really important thing. And it seems to just be increasing every day. We're seeing things, you can actually break the law sometimes if you pay the ransom because where these people operate, there's all kind of weird stuff that's coming out of ransomware as a very specific kind of type of security threat that even elevates kind of business continuity and resiliency on a whole nother level for this one particular risk factor. I wonder if you're seeing some of that as well. It's a great point. In fact, it's clearly an industry that was resilient to a pandemic because we've seen it increase. This is organized crime at this point, right? This isn't the old days of hackers playing around. This is organized crime and it is accelerating. And that's one thing I'm really glad you brought up. It's an area we've been really focused on across our whole portfolio. Of course, IBM Tape offers the best most at the actual real air gapping, physical air gapping. We can take a cartridge offline. But beyond that, we offer you the ability to do, different types of logical air gaps, whether it's to a cloud we support. In fact, we just announced now with Spectrum Protect we have support for Google Cloud. We already supported AWS, Azure, IBM Cloud. So we give you the ability to do logical air gapping off to those different cloud environments. We give you the ability to use worm capability so you can put your backups in a vault that can't be changed. So we give you lots of different ways to do it. And in our high-end enterprise storage, we offer something called safeguarded copy where we'll actually take data offline that can be recovered almost instantly. Something very unique to our storage that gives you for the most mission critical applications the fastest path to recovery. One of the things we've seen is some of our customers have done a great job creating a copy, but when the event actually happens they find it's going to take too long to recover the data and they end up having to pay the ransom anyway. So you really have to think through an end-to-end strategy and we're able to help customers do a kind of health checks of their environment and figure out the right strategy. We have some offerings to help come in and do that for our customers. Shift gears a little bit. We were at AnsibleFest earlier this year and a lot of talk about automation. Obviously Ansible is part of the Red Hat family which is part of the IBM family. But we're seeing more and more conversations about automation, about moving the mundane and the error prone and all the things that we shouldn't be doing as people and letting people do more high value stuff. Wonder if you can talk a little bit about the role of automation, the kind of development of automation and how you're seeing that impact your deployments. Brent, you want to take that one first? Yeah, sure. So the first is when you think about individual Kubernetes clusters, there's a level of automation that's required there. I mean, that's the fundamental, I mean, back to the infrastructure as code, that's inherently, that's automation to effectively declare the state of what you want your application in your cluster to be. And that's the essence of Kubernetes. You declare what the state is and then you pass that declaration to Kubernetes and it makes it so. So there's the Kubernetes level in automation, but then there's what happens for larger enterprises when you have tens or hundreds of Kubernetes clusters. So that's an area of, Jeff, you mentioned Ansible. Now that's an area with the work the Red Hat's doing in the community for multi-cluster management, actually in the community and together with IBM for automating the management of multiple clusters. And last thing I'll touch on here is that's particularly important as you go to the edge. I mean, this is all well and good when you're talking about, you know, safe raised floor data center environments. But what happens when your tens or hundreds or even thousands of Kubernetes clusters are running in an oil field somewhere? Automation becomes not only nice to have, but it's fundamental to the operation. Right. Yeah, but Brent, let me just add on to that real quick. You know, it's funny because actually in this COVID era, you're starting to see that same requirement in the data center, in the core data center. In fact, I would say that because there's less bodies now in the data center and more people working remotely, the automation in need for automations actually accelerating as well. So I think what you said is actually true for the core data center now as well. Right. So I want to give you guys the last word before we close the segment. I'm going to start with you, Brent. Really from a perspective of big data and you've been involved again in big data for a long time, as you look back at kind of the data warehouse era and then we had kind of this whole rage with the Hadoop era and you know, we just continue to get more and more sophisticated with big data processes and applications. But at the end of the day, it's still about getting the right data to the right person at the right time to do something about it. I wonder if you can kind of reflect over that journey and where we are now in terms of this mission of getting the right data to the right person at the right time so they can make the right decision. I think I'll close with accessibility. That's these days as we, you know, the data scientist and data engineers that we work with. The key problem that they have is accessibility and sharing of data. I mean, this has been wonderfully manifest. In fact, we did some work with the province of Ontario. You could look this up, hashtag how's my flattening. So the work with them to get a pool of data scientists in the community, in the province of Ontario, Canada to work together to understand how to track COVID cases so that government could make intelligent responses and policy based on the facts. So that need highlights the accessibility that's required from today's, you know, yesteryear it was maybe smaller groups of individual data scientists working in silos. Now it's people across industry as manifest by that that need accessibility as well as agility. They need to be able to spin up an environment that will allow them to, in this case, to develop and deploy inference models using shared data sets without going through years of design. So accessibility and back to the acceleration and agility that Sam talked about. So I'll close with those words. That's great. And the consistent with the democratization of two is another word that we hear over and over again in terms of, you know, getting it out of the hands of the data scientists and getting it into the hands of the people who are making frontline businesses every day. And Sam, for you, for your clothes, I'd love for you to reflect on kind of the changing environment in terms of your requirements for the types of workloads that you now are, you know, looking to support. So it's not just taking care of the data center and relatively straightforward stuff, but you've got hybrid, you've got multicloud, not to mention all the media, the developments in the media between tape and obviously flash spinning and spinning drives. But, you know, really we've seen this huge thing with flash, but now with cloud and the increased kind of automation of units to be able to apply big batches and small batches to particular workloads across all these different requirements. One of you can just share a little bit about how you guys are thinking about, you know, modernizing storage and moving storage forward. What are some of your priorities? What are you looking forward to to be able to deliver, you know, basically the stuff underneath all these other applications. I mean applications basically is data with a UI and some computer on top. You guys were sitting underneath the whole package. Yeah, you know, first of all, you know, back to what Brent was saying, data can be the most valuable asset of an enterprise. You can give an enterprise an incredible competitive advantage as an incumbent if you can take advantage of that data using modern analytics and AONI. So it can be your greatest asset and it can also be the biggest inhibitor to digital transformation if you don't figure out how to build a new type of modern infrastructure to support access to that data and support these new deployment models of your application. So you have to think that through and that's not just for your big data, which the big data of course is extremely important in growing at an incredible pace all this infrastructure data. You also have to think about your mission critical applications. We see a lot of people going through their transformation and modernization of SAP with move to S4 HANA. They have to think about how that fits into a multi-cloud environment. They need to think about the life cycle of their data as they go into these new modern environments. And yes, TAPE is still a very vibrant part of that deployment. So what we're working on and IBM has always been a leader in software-defined storage. We have an incredible portfolio of capabilities. We're working on modernizing that software to help you automate your infrastructure, ensure you can deliver enterprise class SLAs. There's no, nobody's gonna alleviate the requirements of having near perfect availability. You don't because you're moving into a Kubernetes environment get a break on your downtime. So we're able to give that real enterprise class support for doing that. One of the things we just announced at the end of October was we've containerized our spectrum scale client allowing you now to automate the deployment of your cluster file system through Kubernetes. So you'll see more and more of that. We're offering you leading modern data protection for Kubernetes will be the first to integrate with OCP and OpenShift Container Storage for data protection and our Flash System family will continue to be on the leading edge of the curve around Ansible Automation CSI integration with Kubernetes. So we'll continue to focus on that and ensure that you can take advantage of our world-class storage products in your new modern environment. And of course, giving you that portability between on-prem and any cloud that you choose to run. Exciting times, no shortage of job security for you gentlemen, that's for sure. All right, well Brent, Sam, thanks for taking a few minutes and is great to catch up. And again, congratulations on the success. Thank you. Thank you. Thank you. All right, he's Sam, he's Brent. I'm Jeff, you're watching theCUBE's continuing coverage of KubeCon Cloud NativeCon North America 2020. Thanks for watching. We'll see you next time.