 Hey, welcome back to Los Angeles. Lisa Martin here with theCUBE. We are live at KubeCon and CloudNativeCon 21. It's been great to be here. We've been broadcasting the last couple of days, about 2,700 people joining us in person. Great buzz, great energy. I've got two guests here next joining me remotely. Please welcome Itzik Reich, the VP Technologist at Dell EMC, and Nivis Ayur, Senior Principal Product Manager at Dell Technologies. Gentlemen, welcome to the program. Thanks for having us, Lisa. Thank you, Lisa. And we're pleased that you're joining us today. Itzik, let's go ahead and start with you. Let's talk, we've seen a lot of uptick and Kubernetes has been picking up a lot. What are some of the things that you're seeing through your lens? Right, that's a great question, Lisa. So really we need to take a step back probably into 2019, we just mentioned in-person conferences. So back then we started to see a slow adoption of customers that are starting to play with Kubernetes in their test environment, maybe running some POCs, but then the pandemic happened obviously, and we started to see huge explosion in terms of adoption and accelerating their digital-based projects for our customers. So they're really starting to pick up Kubernetes and use it heavily in their production, of course, in addition to their test and dev environments as well. And because of that adoption, they started to think about other scenarios and other considerations that are relevant for their production environment, which is based upon Kubernetes. Things like disaster recovery, availability, all of those things that typically you don't worry about when you just run them in a small desk or a POC environment, but are super critical for our customers. And as the largest storage company in the world, we have the smallest customers in the world, but also the largest and the most demanding one. So really huge adoption that needs to basically accelerate all of those aspects that belong to an enterprise environment that happens to one on Kubernetes itself. You've asked, do you see something similar? Yeah, absolutely. I agree with Idzik and actually one of the brief stories actually I start up with is because a few years ago actually, several years ago when I was taking a cab in New York, remember the point of sale terminal was not working. So you took my credit card, it's just like used to magnetic drive. So not having the technology access was like an inconvenience, but it still could transact. But now today's age, when you look at digital transform, digitally transform companies, starting with all these web companies, like Uber, Lyft and things like that, but then you also have mainstream companies where the entire business is now taking over digital. And so all these applications are the ones that are powering the entire business, if you will. And not having these applications available or these apps available will basically the business going to lose money. And that's what is, and the pandemic has only accelerated digital transformation because everyone working from home and also the customers are also remote. So now you have the entire operation is just softwares running the business. Pretty much every company is the technology company. And then you have, and then all these applications, they are modernized. So they are modernized in the way that they're not built to the traditional architectures. They're now using microservices, DevOps and Agile. These are three major aspects that kind of drove the new modernized applications to build more complex applications. And Kubernetes has emerged as a sole platform that can serve the underlying platform between all of these aspects. And hence we see that Kubernetes adoption has taken off a lot because pretty much every organization is running several projects within the enterprise, including app modernization, you know, transformation of any kind of secondary kind of use cases, IOT, you know, the whole digital transformation story is kind of running on Kubernetes. And as it was pointing out, so now Kubernetes has emerged as the key infrastructure as a service layer, if you will, about infrastructure service. And it needs to consume storage and it needs to have, you know, all these traditional capabilities that were for applications, right? I mean, like disaster recovery, having enterprise-grade availability aspects, like, you know, for data protection and things like that. And that's sort of, the enterprise capabilities are relatively, I would say, accelerating a lot. Earlier Kubernetes was more on the non-enterprise aspects of the journey. Now we are seeing a lot more enterprise. Are you seeing your conversations within organizations elevate up the chain where Kubernetes is concerned? Is this a C-level conversation? Are they understanding that from a competitive differentiation perspective, from a modernization perspective, it's the direction they need to go in? Yeah, absolutely. For them, you know, VM will run itself a couple of months ago about the reasons that are important for customers to run containers in production. There were like, 10s of them. But the number one reason is to accelerate software adoption. Basically, write codes faster. That's like the number one reason. It's not about the technology itself, you know, technology is just an enabler. And the enabler is to write the code as quickly as you could, deploy it in test and dev quickly as you could, run some QA cycles on it, and release, release, release the code. That's at the end of the day. That's the main difference between the old way of the waterfall approach to the new way of agile approach, which eventually gets translated into the infrastructure layer itself that needs to accommodate those changes, if you will. Well, releasing code faster is going to enable organizations, presumably in any industry, to be able to develop and release products and services faster to the demanding consumer market, I imagine. That's absolutely correct. We've all got spoiled by the smartphone industry or just expect a new version to be just deployed to your device almost every day now. It's exactly the same thing. It is, I think we carry that. I think it's impossible not to carry that consumer expectation from our consumer life into our business life. And we just expect that things are going to work that way because in our consumer lives, they do. When I ask you guys about, is this question is directly for you, talk to me about CSI. What is it besides a TV show? I know you have a great answer for this. And many spin-offs, by the way. Right. It's not just a single one. CSI, right. So let's take a step back in 2015. Back at DockerCon, they sit on the stage and during the keynote, and they explained that you should write your code in the 12-factor way. Resiliency should be built into the containers themselves and you shouldn't care about storage persistency. Now being the storage industry for the best part of my life right now and storage persistency is important. If a customer lose data, that's a very bad day for the customer and possibly a very bad day for me as well. So it's all about the data. Nothing else really matters. The data itself is the goal. And so there was no data persistency back then in Docker. And we actually work with a startup that did just storage persistency for containers, basically meaning the ability to provision a volume from storage array into Kubernetes and Kubernetes will know about this. That startup went busted, but the need still existed. And so into that need, Google came and they come up with this API called Container Storage Interface short for CSI. And it does exactly that. It allows Kubernetes administrator to Kubernetes API to consume storage from the underlying storage arrays. So provision volumes, mapping volumes, taking a snapshot of the volume and mapping those very basic capabilities. Now those capabilities are very basic. And we now have customers that are telling us I need far more than just the ability to provision a volume for my Kubernetes environments. I need this volume to be protected. I need this volume to be replicated. I need this volume to be protected into a backup device. All of those things that CSI doesn't know to do today, no, we didn't know to do in the near future. So what we did is we said, right, we're not going to reinvent the wheel, that's CSI. We're not just going to repeat CSI all over again. We're going to extend CSI with open source tools that will enable our customers to do all of those things that I just mentioned before. So CSM is Container Storage Modules, which is what we announced today. And it's very high level, provides you the capabilities to do the following. The first one is the Observability Module. So if you're monitoring your open source environments, you are very, very likely to use open source tools like Grafana and Prometheus. So we have this plugin that allows you to monitor your storage array with Grafana and Prometheus and really becomes a liaison point between the storage admin, the Kubernetes admin. They can connect both to the console and each really understand the entity that he is not aware of. I call it the two-way mirror base. Second module is the Resiliency Module. Kubernetes is very infant in terms of understanding storage. It doesn't understand storage fellow conditions. And so our Resiliency Module run as a K3S. It's like a mini-me version of Kubernetes, if you will, which keeps monitoring both the storage array and the host. And in case of a storage error, it knows to act upon it and do things like volume unmapping and map those volumes to other surviving servers in the pod, et cetera, et cetera. The other module is the Replication Module. So back into 2015, customers are basically telling us today, I want to use Kubernetes, but I also want to replicate the data to either a passive site or an active site. And in case of a failure, if my primary site goes down, I want to fail over these Kubernetes volumes and data to a remote site. So literally within a click of a button, you can fail over your Kubernetes environments from site A to site B using the underlying storage array capabilities, the application, et cetera, et cetera. And the other module that we've also announced is the Volume Group Snapshots. So instead of just taking a single volume, which is what CSR is all about, you can actually take multiple volume that belong to multiple microservices at the end of the day running within those containers in order to really back up a service and not just the microservice itself. So all of these modules and future modules that will come in the future as well, belong to CSM and CSM for us is just the beginning. It's everything that most demanding customers want us to provide today and they are not willing to wait for CSI to catch up, please. Got it. So you've done a great job of explaining what CSI is, what it isn't, what CSM is, and all the great things that were announced today. Let's talk about the data protection, the security angle. We've seen so much change in the security, the threat landscape in the last 18 months. We've seen ransomware become a household word, the proliferation of DDoS attacks, and of course there's this scattered workforce that is still scattered. Talk to me about why data protection for Kubernetes and what are some of the unique needs that that presents? Sure, thanks, Lisa. So when you look at the Kubernetes landscape, it's originally started out with mostly the front end aspects, mostly like web tier type applications, but as the landscape has evolved, now we are seeing actually in the Kubernetes community also there has been newer concepts, like stateful sets, for example, which allows you to have more persistent type or basically the application that have retained state and data in the Kubernetes cluster. And we are seeing a huge proliferation and that is also increasing across the board on, for example, everything from experimentation or like any kind of user experience, kind of data understanding about sessions, what users like, what they don't like, to all critical operational aspects, to transactional elements too, all of them being brought into the Kubernetes. We are seeing organizations in various stages of the journey and then in add on to the additional capabilities on the storage side, as it was mentioning about CSI and CSM, and basically the ways for the Kubernetes layer to consume these storage services. So when you're building these modern applications, the state is now preserved as part of the Kubernetes. And actually recently we had a case with one of the customers we've had and so they did not have data protection as part of the Kubernetes. And we are seeing this in several organizations where you have an IT ops kind of a team and there is a DevOps team, there's a two-speed IT concept. So DevOps teams, a lot of time, they do not take into consideration a lot of these disaster recovery and data protection aspects as part of the design. And then one of the customers just what happened and they lost data because there were systems crashed and it was not through ransomware, luckily, but it was through a general logical failure of hardware, things like that. And so they could not recover that. So they had to go back and they had to reach all the whole thing. So they started investing in saying, oh, we need a ways to protect the data so that I can recover. So data is all about recoveries, about making sure you can recover to a certain point in time and also recovering in the minimal amount of time. And the challenges that Kubernetes adds on top of traditional application that the entire application definition in Kubernetes is split across multiple of these smaller metadata aspects, the application itself will have labels, they will have, you know, they'll have secrets, they'll have config maps, they'll have custom resource definitions. They have all this additional metadata that make up the entire application, not just the data. So you need to have all of that captured in context in a cloud native fashion if you're trying to protect that Kubernetes environment. And that's kind of a little bit of a unique and challenge. And then from a security aspect that you alluded to from Kubernetes, there have been multiple security challenges that we see, although we don't directly work on the front end on the prevention side, but on the cure side data protection is an important aspect, right? I mean, if you look at the open source community, there is so much open source today, and how do you know that the open source and the API that you're consuming is coming from a valid source? So you need, so there is all kinds of like different security solutions that Kubernetes community offers to validate, making sure the source code is good, the APIs are authenticated, and you know, authorize things like that. So there is a lot of these, but even despite that, you know, there is always ability for some attacks to sneak in. And that's where data protection is providing that cure. So in case something does happen and you have a ransomware attack, I have a cybersecurity vault or I have data stored in a secure fashion in a panic room, if you will, that I can, so my business, like I was alluding to my earlier example, the business depends on that data and that operational transactional customer data, and you need to recover that, and you need to store it in a secure place. And that's sort of the whole aspect of that. It's got to be recoverable, that's the whole point. Guys, thank you so much for joining me, talking to me about what you're seeing from a Kubernetes adoption acceleration perspective. Thank you for helping me learn a new definition of CSI, not a show or a spinoff, and talk to us about what CSI is and the things that you are, the modules that you're announcing today. We appreciate your candor and your time. Thank you, Lisa. Thanks for having us here. Thank you, Lisa. I appreciate it. My pleasure. For my guests, I'm Lisa Martin, coming to you live from Los Angeles at KubeCon CloudAid of Con 21. We'll be right back. Dave Nicholson will rejoin me with our next guest. Stay tuned.