>> Narrator: From Burlingame, California, it's the Cube, covering Sumo Logic Illuminate 2019. Brought to you by Sumo Logic. >> Hey welcome back everybody! Jeff Frick here with the Cube, we're at the Sumo Logic Illuminate conference, it's at the Hyatt San Francisco Airport. About 700, 800 people, full house in the keynote earlier today, all about operational process monitoring, all this crazy data is being kicked out of the Cloud and IoT and all these crazy next-gen applications. We're excited to have a very close friend of mine, CEO of a very hot company, Kamal Shah, the CEO of StackRox. Kamal, great to see you! >> Thank you, and great to be here, Jeff! >> Absolutely! So for folks that aren't familiar with StackRox, give us the overview. >> Sure, so in a nutshell, we do Kubernetes Security, and so as we've heard all day today, enterprises are deploying microservices, containers, Kubernetes, and we do security for your cloud data infrastructure. >> So how does security work for Kubernetes versus security for other things? >> Yeah, so the use cases for security, or the mission for the security team is the same, right? You got to harden your environment to prevent the bad guys from getting in. >> And, you have to make sure, despite your best efforts, if somebody does break in, then you catch them before they do any damage, right? But the how you do security has to evolve for the cloud data stack, right? It has to understand the containers are immutable affirm all infrastructure, you have to understand that it's not just about the container, but it's also about the orchestrator, and specifically Kubernetes, and it's also about making sure that you seamlessly integrate with dev ops processes, automation and workflow. So it requires a fundamentally different approach to security than traditional security tools. >> So you know, we talk a lot about the increasing attack area that's offered by IoT, right? And increasing attack area that's offered by all those API's and all these interconnected applications, but I've never heard anyone really talk about containers or orchestration as kind of a new attack surface. Did we just stop paying attention? Is that something you're seeing happen? >> Yeah it's something that is starting to emerge, and we've seen some high-profile breachers at a large next generation electric car company, and a large shopping site where misconfigurations led to security breaches in the Kubernetes' environment, and Kubernetes' ecosystem also did a Cube security audit, and so I think we're going to start to hear a lot more, because there's more and more applications are being deployed in production. It's creating a new attack area, and as the old saying goes, the predators go where there's food in the system. >> And so if you're not proactive about it, I think it's going to really hurt as you deploy containers in Kubernetes. >> Right, so we hear over and over and over again about breaches because people misconfigure stuff. That just seems to happen, whether it's a database or this, that, and the other. And I think we can pretty much safely assume everyone's going to get breached if they haven't got breached already, 'cause we hear about it all the time. How do you catch them fast, limit the damage and try not to have too much vulnerabilities? >> Exactly, so the use cases for what we do at Kubernetes are the same. Right? Its vulnerability management, it's configuration management, and we just did a study around the state of container in Kubernetes security and misconfigeration was the number one concern. Because the reality is that Kubernetes, there are a lot of knobs. And each knob has multiple options, so if you're not careful you can really misconfigure your environment and make it so much easier for attackers. >> Right, right. >> And it's precisely what happened at the two examples I sighted earlier. So a misconfigerations is important, runtime security is important, and also compliance. Let's not forget about compliance, right. You have to make sure that you meet your PCI, HIPAA, NIST, and CIS benchmark standards for this cloud native stock. >> So what we're seeing is that these are all becoming very, very important and as a result, it's increasing awareness as Kubernetes gets more prominent. >> Right, and then they are creating and tearing down hundreds, thousands, millions of these things at a nidicolous pace. >> I mean exactly. Kubernetes came out of Google, they open sourced it, and it's really what allows you to deploy, manage, containers at scale. Apparently, they manage hundreds of millions of container a day using Kubernetes, it's incredible. >> Jeff: Oh yeah, I saw a statistic that Google launches 4 billion containers per week. That was from a presentation, actually from a 451 analyst from like 2 years ago. So one can only imagine the scale. >> We are also seeing not quite 4 billion containers per week, but we are seeing thousands, and tens of thousands of containers at scale at companies everywhere. They are all deployed in production, and now they are waking up to security. The good news here is that they are waiting for breaches to happen before they solve the problem. There's still a lack of awareness, and what Sumo Logic has done today with the announcement around continued intelligence for Kubernetes just increases the awareness around, hey we have to solve observability, which is logs, metrics, and tracing, which is what Sumo does, and security for your cloud native infrastructures. >> Yeah, I mean the automation is so important, right? You can't do any of this stuff with this exponential growth of data, exponential growth of pushes, of new code releases. There's so many pieces in this, so automation is a huge piece of the puzzle. >> Automation is paramount and with this new infrastructure there aren't enough security people to solve this. So security has to become everybody's responsibility. And the only way we are going to solve this is to automate it. It also has to integrate with your DebOps processes and automation and work flows. If you don't, then the DebOps body is going to reject the security organ, right? So it has to be seamless in the way you deploy it. >> It's interesting you say that because we go to RSA, forty thousand people, more vendor than you can count, it bulges Moscone to the absolute edges. Everyone says over and over that security has to be baked in the entire process from beginning to end, it's not a bolt on and can never be successful as a bolt on. So it surprises me to hear you say that still a lot of people are kind of behind the curve. >> Well I mean if you think about I, even though they say that, right? In a traditional model of the application you go to spend 6 months building it and then you can go spend a couple of weeks or month hardening and putting security around it. But when you are launching applications every 6 hours, you can spend 6 days addressing security, so it has to be built in. And speaking of RSA, if you recall, last year the big talk at RSA was around AI, right. Everything was AI driven security. My prediction, my bold prediction for this RSA is it's going to be all around Kubernetes security. >> Yeah, well it's applied AI. Applied AI for Kubernetes. >> Exactly. >> And that's what you need. I always feel for the SISO just walking the floor at RSA going, "Where do I begin? I mean where do I spend my money, how do I prioritize?" It's kind of like an insurance problem. You can't insure to the nth degree. You got to have a budget, but how do you deploy your assets? It's got to be super, super confusing. >> It really is. I think what your seeing is that SISO's are relying on their DEV and IT ops teams, right? They are partnering with the VP of platform, the VP of infrastructure, the VP engineering, because when you think about this new world security is really, the ownership of security is now shifting from the information's security teams to DevOps teams. So security teams still drive policy, and they still want to make sure they do the trust and verify, but the implementation of the security is now being owned by DevOps teams. So its a big cultural shift that's going on in organizations today. SISO's have to realize that it's no longer just them, but they have to partner with their DevOps counterparts to effectively address security for this cloud native stock. >> Right, so tell us a little bit about the relationship with Sumo. How do the applications work together? What's the solution look like when the 2 solutions are brought together. >> So Sumo has been a great partner. We have several joint customers. The simplest way to think about this is that Sumo does observability for Kubernetes, so that's logs, metrics, and tracing, and we do security from Kubernetes. We are the yin to their yang. What we do is we have taken all the intelligence we get from security and we feed it into the Sumo dashboard. Sumo customers get a single pane of glass, not just for the observability data, but also for their security violations, weather its for vulnerability, weathers it's for configuration or if it's for runtime threats, right? You get it all in one single place. >> Right. So I just want to get your take on kind of this rise of the momentum behind Hybrid Cloud that we've seen recently. Big announcement at Google Cloud show, with Anthos. Big announcement between VMware and Amazon. It always kind of swings back and forth. It was all in to public cloud and now there's a little bit of a pullback in Hybrid, but that's terrific for you. The fact of the matter is workload should run where they should run, they don't really care it's what's appropriate. Horses for courses, right? >> Precisely so, we see the shift from public cloud to Multi-cloud, and then from Multi-cloud to Hybrid cloud. The underlying infrastructure that makes that a reality are containers and Kubernetes, right? And that's why we've seen this tremendous momentum on Kubernetes. What we are seeing is customers that want to give their Dev teams that flexibility to pick their favorite cloud, or to do it on premises, their private clouds. But they want to make it in a single security solution that gets integrated no matter where you run your infrastructure and that's integrated back to your Sumo dashboard. So you have visibility across all Dev teams, all your application infrastructure, regardless of where they are running. There is one security standard that gets implemented. That is really, that's the future. You don't want to be beholden to a one claw provider, you want flexibility, you want choice. Kubernetes allows you to do that. >> Well and the whole thing becomes more autotomized, right, with autonomic memory, autonomic compute, autonomic store, throw that on an IoT and Edges and now you're starting to distribute all those pieces all over the place, which is going to happen. >> Kamal: It is going to happen for sure. >> All right, looking forward I can't believe we're almost through 2019, it still shocks me everyday I look at the calendar, but what are some of your priorities looking forward? What are you guys working on? What do you see coming down the pipe? >> Yes, so you touches on a couple of these. So today, is a lot of talk around Kubernete. We are seeing Kubernetes also get deployed in IoT and edge devices, we are also seeing they are being used to manage serve-less infrastructure. So we are going to continue to evolve as Kubernetes evolves. The other big trend that we are seeing in the market today is around service mesh. People talk a lot about Istio and Linkerd and using service mesh as your policy framework to drive consistent policies across applications, so that's another area where we are innovating very rapidly and that will become, I think, more and more real in enterprise deployments over 2020. >> Well, congratulations Kamal to you and the team. I think you picked a good horse to ride on, I should say ship, right, with Kubernetes. Thanks for taking a few minutes. >> No, thank you for having me. I can officially say now that I've checked off one of my professional bucket-list items, which is to be on the Cube with an old friend. So thank you for having me. >> Check that box man. All right, he's Kamal, I'm Jeff, you're watching the Cube. Were at Sumo Logic Illuminate from the Hyatt San Francisco Airport. Thanks for watching, see you next time.

>> Announcer: Live from Washington D.C., it's theCUBE covering AWS Public Sector Summit 2017. Brought to you by Amazon Web Services and its partner, Ecosystem. >> Welcome back to our nation's capitol where we continue our coverage here on theCUBE of the AWS Public Sector Summit 2017. Some 10,000 strong in attendance this week here in the Walter Washington Convention Center. It's just about a mile from the U.S. Capitol. John Walsh, this is John Furrier. John, do you feel the energy of the centerpiece of the political universe. >> It's hot here in D.C. >> It is hot. >> It's a pressure cooker, the humidity. >> But, it's not global warming we know that because, ya know, climate change is >> Climate change is not real. That's from what I heard. >> That's what we've been told. >> The problem with D.C. is it's a data lake that's turned into a data swamp. So, someone really needs to drain that data swamp. >> Well, ya know, to help us do that. You know who's going to help us do that? >> Amazon Web Services. >> Marlin McFate's going to help us do that. He is the technical leader of the Advanced Technology Group in the office of the CTO and Riverbed. And Marlin, thank you for being with us here on theCUBE. Your first time, I believe. >> Yes, it is my first time on theCUBE. >> So, you're a Cube rookie? >> Yes, Cube rookie. >> Good to have you aboard. >> I appreciate it, thanks. >> Tell us a little bit first about Riverbed, about what you do there specifically, what you do there and what the company's mission is overall. >> Absolutely, so I work for the Advanced Technology Group, the Advanced Technology Group works underneath the office of the CTO. There's actually two groups that work under the office of the CTO, my group, the Advanced Technology Group and another one called the Strategic Technology Group. The ATT Group, the one that I belong to, we focus on being the subject matter experts of our products. I think there's about nine of us now and we all focus on different products. Riverbed's grown from a company of being just the WAN Optimization Company to really being the performance company, right, whether that be visibility, whether it be optimization, whether it be network optimization. Each one of us focuses on a different piece. I, predominantly focus on our WAN optimization, our SteelConnect product and at times our SteelFusion project, which is the combined Edge product. >> SteelConnect, yeah, tell us what that's all about. >> SteelConnect, SteelConnect is not actually our most recent product to come to market. We have a couple of visibility products that have come out recently, but SteelConnect addresses the idea that we have been doing networking for the same way say, you know, 1993 beyond, right. We are still doing it the same way. Everything within our industry, whether you take a look at virtualization, whether you take a look at Cloud, whether or not you take a look at storage, everything has changed substantially in how we do it and this brings that change to networking. The idea is that when you think about servers you say, I no longer want to think about you know, hardware. I never want to think about that. I never want to think about resources. Maybe I don't even want to worry about operating systems. I only want to worry about containers, right. Now, when it comes to networking I don't necessarily want to have to worry about each individual piece within my network. I want it to be orchestrated and controlled centrally and what I tell it to do, I want it to do. I shouldn't have to do that. >> You missed a challenged. We heard Vernon Vogel on stage here at Amazon a couple of seconds ago say, I'm here in D.C. say hey, it's a new normal. We had another entrepreneur on just before you from FUGE who said, hey, it's inevitably the world of the future and it's inherently different, or intrinsically different in Cloud than it is on premise with enterprises, so the question for you is, what is the use case that you guys are winning at because the Cloud is impacting federal government and public sector, but a lot of times they have old, antiquated systems like back in 1993, '94. So, they're moving fast to commercialize, to modernize, that's the focus. How do you guys help them? What's the big lynch pen for you guys and that goal mission to the customer? >> Alright, so you're absolutely right. The government has been here, or the government or public sector as a whole has been moving to the Cloud quite quickly here recently, right. We've seen this move more on the commercial side first, obviously, and now in the public sector. One of the very large use cases that we address is the ability to provision for your applications, right. Some of the characteristics that you find in commercial world, such as, I want to use internet as transport. You don't see as much in public sector. But, you do see, I can spin up an application in the Cloud. If you go to your Cloud person and say, how would it take me to get application B, they could possibly come back to you and say, well, would this afternoon be okay, right. Can you provision in hours like that? Can you get the policy in place for users? Could you get the connectivity? Could you get any of that in place in the same amount of time? That is a use case that SD WAN addresses without having to rip up, take out the network that you already have, which is the physical network, or what we refer to as the underlay. Being able to give you that flexibility on top of that network. >> The big thing that customers have a challenge on is that other focus it's DebOps trend programmable infrastructure is another one, so that they want to make it programmable. >> Right. >> So, how do you guys fit into that? Because one of the things that we hear is, could I have develop 'cause all I want to do is have infrastructure just works as code. That's all I need for whatever use case. >> Yeah, we usually see that DebOps is actually one that'll probably be the first movers to the Cloud for the public sector, right. With our, really it's every single one of our products, whether or not we're talking about SteelConnect, SteelHead, SteelEssential, any one of them, there's a RESTful API for every single one of them. So, you can actually go in and utilizing a very easy scripting a RESTful API directly itself and spin up whole environments and then spin them down if you wanted to do that. So, it fits very, very nicely into that DebOps world. >> Do you have SteelEdge yet? >> SteelEdge? >> Copyright on theCUBE. >> It might be a razor company that might have that. I don't know. >> Well, the Edge in the network is huge and this is where we're talking about as you guys do it, you know SD WAN, I mean, come on, why the area networks? You don't beat, you can't get any more edgier than that. You guys have a core competency in this. How do you guys look at the Edge and IOT and all these use cases popping around? >> Well, we do actually have a product that has Edge in it, it was SteelFusion Edge. We could address that in a couple of different ways. I want to make sure that I understood your question, though. Your question was around IOT, specifically? >> Well, how do you guys look at the Edge? The trends right now are super hyped up right now, Intelligent Edge is a big message we're hearing from others. IOT is an Edge application with its Industrial Edge with Genery Censor networks, help with safety, surveillance, all this is Edge devices. >> It still ends up in the end being you know, and that has, we've heard the change from people calling it Branch to calling it Edge, which is probably pretty appropo, right. But, really in the end, what it comes down to is connectivity, right. So, if I have IOT sensors in a warehouse, whether or not I have an application, whether or not I have a group of users, whether or not I have mobile users, in the end what it really comes down to is connectivity. And, we all especially with our cell phones, right, we have come pretty much to the point where we expect our data and our connectivity to be there at all times, right. That's one of the things SD WAN addresses. Whether it be our direct, our SD WAN products, SteelConnect, or whether or not it has works with some of the pieces that move further into the LAN architectural, like our wireless access points, our switching, right. So, you can imagine here, right, I can provide policy for my IOT devices. I can provide that policy one time at an organizational or agency level. I can have that policy filter down, all the way down to the axis point and now the axis point might be my axis point to my IOT or to my user. So, in the end, it still comes to connectivity. >> Marlin, what's some of the use cases or scenarios you've been involved with customers where it's been super exciting from an architectural standpoint, where you guys are doing some cutting edge things. Like, is it more the network size? Is it software? Is it Edge. I mean, I'm tryin' to get a sense of, could you share a personal perspective? >> Absolutely so. One of the ones that we're working on right now I think is probably the most exciting. It is combining some aspects, you could call it an FE. You could call it SD WAN. You could call it Grey Box. What I like to call it is just a combined Edge piece, right, which encompasses both the SteelConnect piece which handles your firewall characteristics, your identity management characteristics, built into that some switching, virtualization, so you can run other products on there. What the customer really wanted to end up doing was they had school systems that, a school system that was in a very far away place and that school system, they were putting in a router, a switch, an access point, you know, all these different little pieces and devices, right. What we did was we were able to take that design and crunch it down into basically one box, right. They have enough switchboards. They have the ability to run virtual machines 'cause they said that they had a server here or there. They have their virtualized SteelConnect gateway which gives them the firewall capability, gives them the routing capability and this is all combined in a box that already has the WAN Optimization built in. So, they get everything that they would have had onsite in one box. >> Is there something to working, you bring up education as an example, but in that space overall in the .gov, the .edu space that's separate and aside from commercial partners or commercial relationships like different concerns, different priorities and yet they're using the same technologies. >> Most certainly. The only thing that I could really say from a using technology, right, I mean there are some pockets where different technology, far off weird technologies is utilized. But, I would say that they are the public sector, schools, federal government, intel, they're all using a lot of the same technology, right. It's when they adopt it. When did they bring it into their environment? And then, what are the special characteristics of their environment? So for example, what I said earlier, right, your commercial customers are looking at utilizing SD WAN to move maybe completely off of MPLS. It's probably not something that we're going to see within the public sector, right. They're want to still use some sort of private networking. I do have some customers that are utilizing public internet, but then, they are tunneling an overlay back to an MPLS entry point to get back into their Cloud. We just have interesting requirements. Whether that be a trusted internet connection, whether or not that'd be JRSS, we have different security requirements in the public sector. >> Well, I love some of what you're doin'. Did you get all of that MPLS stuff there? >> Yeah, I got the first four. >> I want to jump in and double down on that. This is interesting conversation because the whole trend right now is hybrid Cloud on the Enterprise side which is a leading indicator to the government, a little bit lagging on that, so whatever that translates to in terms of Hybrid or Legacy, it's going to be somewhat similar, I believe. But, really multi-Cloud is a trend that people are talking about. It's super hyped up but it's not yet real. The thing that's holding multi-Cloud back not multi-Cloud in the sense I got to workload over hear and a workload over there, I'm talking about moving resources around the network, data, compute, what not, is latency, huge problem. You mentioned MPLS and all this tunneling, there's still the latency problem of how do you get the laws of physics down to the point where you can actually have those kinds of latencies? What is Riverbed doing? Can you share some insights to that direction 'cause that's the holy grail right now. That's the last hurdle. Then, well getting all the silicons is still the final hurdle, but latency's critical. >> So, problem number one there, right. Even if it is Cloud to Cloud in that example, right, is first how do I get a WAN Optimization device, something that can optimize that traffic for me. Something that can affect my latency for me into that environment. Riverbed has worked tirelessly to get that in there right. But, to your point, you can't change how an electron flies, right. The speed of light is the speed of light. You're not going to get an electron to move any faster. So, what Riverbed developed that's still very relevant today is the ability to, instead of change your latency, mitigate the negative affects of your latency, right. So, if I. >> Or work around it. >> Absolutely, and you can do that at the application level, absolutely, program around it, but there are a lot of protocols out there that aren't necessarily optimized for that longer latency environment, right. So, what we do is, or the adage is, the trip never taken, right, the shortest trip. So, if I have to, not to get into the weeds or anything like that, but if I have to make a thousand round trips to accomplish something, right, and I could put something in there that understands what I was getting, right, that data that I was getting each one of those times and I can take less trips, well then, that just made that faster. So, if I have a thousand round trips and it takes a minute to do, and now I can do ten round trips and it only took ten seconds, or six seconds if we're doing the math right. >> It's kind of like here in D.C., you're local. I noticed that coming from Dulles Airport they have Sirius pricing on the toll roads. That's basically private networking right there. >> That's right. >> These cost path routing opposed to the other side. I was in the, you know. >> Marlin was more describing my trips to the hardware store on the weekends, a thousand round trips, be a lot more economical. But you're right, it is private networking. >> If you're off the road, you're off the packets aren't on the network it saves some room for someone else. >> More traffic, you hear more traffic at the higher speeds. >> You actually could. So, you get two benefits. One is the increase of speed, but the other is the perceived capacity increase of your network. And, we accomplish these things through compression which is really, really simple. I think compression is a must, right. But, through our data duplication. Data duplication is I've seen these patterns before and it's a byte level. We're not talking about an object. I haven't seen a file. No, I've seen these byte level patterns before, I don't need to resend them. And, in traditional network or traditional applications you see pretty much in any organization, right, you typically can get somewhere between 50 and 80, if not sometimes 90% reduction total in traffic. >> My final question before we wrap up this segment here is, Share with the folks, take a minute to talk to the audience about what you're doing with Riverbed at the show and what they should know about the current Riverbed. I know you've guys trying a transformation of yourselves, give a quick plug. Go ahead. >> Absolutely. So, what we're specifically doing here or one of the pieces that is a differentiator for us and our SD WAN is, we went ahead and we thought why couldn't we make that an AWSPPC or a Cloud instance one of my Edge sites, right, connecting into the Cloud, there's many different ways to do it, but why couldn't we make a very simple way of doing that? Why couldn't I take the technology that I'm already putting in place at my data centers, I'm already putting in place at my branch offices, why can't I utilize that to create a secure connection into my BBCs. And, to your point, actually earlier one of the things that's also interesting was Cloud to Cloud. Why couldn't I take that same technology and connect multiple Clouds? Whether they be private Cloud or two public Clouds or connect them all together and take the best of all worlds, right, the best from each and make the best infrastructure that I possibly can. So, what we're showing off here from a SteelConnect perspective is our ability to do that. I can take an AWSVPC, actually I can take all, I think there's 16 regions within AWS and I can interconnect them in less than 10 minutes with the click of a button. And, then back into my infrastructure. So, that and then we also have brought Eternity, which is one of our visibility products that is basically rounded out on our visibility play within the market. We have the network. We have the app. We have the database. Now, we have the end users computer. >> Alright, well, if you could interconnect me to my home in 10 minutes I'm a client. I'd be sold, I'd be all over it. >> I'm going to be in the same traffic as you later. >> I'm not that far from here, but it might as well be another day. Marlin, thanks for the time. >> Absolutely, my pleasure. >> Good to have you on theCUBE, alright. >> Thanks, hope we get to do it again. >> Riverbed has joined us here on theCUBE. We'll be live with more from Washington D.C. right after this.