 Rock and roll. Live from San Francisco, extracting the signal from the noise, it's theCUBE, covering VMworld 2015. Brought to you by VMworld and its ecosystem sponsors. Now your host, Brian Grace Lee. Welcome back, I'm Brian Grace Lee with Wikibon. We're here at VMworld 2015 here in San Francisco, wall-to-wall coverage with Silicon Valley's theCUBE. We've been talking to customers all week. We've been talking to vendors and practitioners all week. Really good vibe all week. But one of the things that we're hearing over and over again, so Pat laid out a fantastic sort of long-range vision, whereas VM, we're going, one of the challenges in the economics and the industry for customers. One of the things that we heard over and over again, though, in the hallway from a lot of practitioners, whether it's from a V-Mug, V-Experts, others, is they say, you know, sometimes all this technology sounds great, the promise of it sounds great, but keeping up with it's really hard. They don't always have time to go to training. You know, change is difficult. And so, today's panel, we're really talking about a couple of really interesting things. First off, yesterday we had a number of the VCs on here. John Furrier was talking to a number of the VCs, really kind of laying the foundation for a lot of XVM where people that had become VCs sort of the new V-Mafia. Well, our panel today, we're excited. We've got a number of startup companies or new companies that are sort of rebuilding the new technology from VM, or outside of VM, where the new V-Mafia. So, I've got Madura Makowski, Makowski, co-founder from Platform 9, Shiv Argawal, co-founder and CEO of Arkan, and John Blumenthal, co-founder, originator at Cloud Physics. Everybody welcome to the show. Thanks. A lot of, VMware's always been known for having great technology. We were talking earlier, all of you were there overlapping somewhere from 2006 to, you know, for a number of years, great technology, you've all sort of gone off and started your own things, you're all building. Talk about, you know, not only introduce yourself, but talk about what your companies do and the types of problems that you're trying to solve that maybe you couldn't have solved just, you know, putting software on premise. Sure, so I'll go first. So I'm Madura Makowski, co-founder and VP of product at Platform 9 Systems. And we started around 2013 or so. I've been at VMware before that for about eight years. And what Platform 9 does in a nutshell is we are OpenStack as a service. So we deliver enterprise-grade OpenStack. But what's unique about our offering is that we package it up and deliver it to our end users as a web service. So it's a SaaS offering. So think of it like your Salesforce or your box. Okay, excellent. Shiv. Thanks, Brian. Thanks for having me here. I am Shiv Agarwal. I am the co-founder and CEO of Arcane. And basically, in the last few years, what has happened is that there's a big shift that just started on the data center side, specifically on networking and security. As you hear about micro segmentation, about virtual networks, SDN. So enterprises have really started adopting those technologies now. And one of the big challenge that they face as they look into these new technologies is it's challenging 20, 30 years of how networking and security was done. And as they look at these new ways of doing networking and security, they are trying to figure out how do they leverage their existing workforce, their existing admins, operators, the existing skill set that they have developed over time to operate these new technologies, right? So VMware and many other data center vendors have thrown out a lot of infrastructure there. What Arcane does is that it simplifies the overall manageability and operations of that infrastructure, which is becoming software defined. Yeah, yeah. And you're helping folks deal with that new potential, but helping them manage across the network, across all that infrastructure. Absolutely. I mean the key is to connect it back to the workload, the compute. So we help understand this new era of networking, new era of security. We map it to how it was done in the classic old days so that we, the existing people, operators with existing skill set can understand it. Then we tie it back to applications that has network and security is changing. What is the impact of it on application? And just as a clarification, we offer both SaaS and on-premise. And because the enterprises who are doing this software defined networks and micro segmentation are large enterprises, and many of them are not today comfortable, and I think that's a topic I'm sure you'll be spending a lot of time on, to send the data to the cloud, right? So the way we have designed it, we enable them to keep the data on-premise. It's fully on-premise model and SaaS, so we give a choice to the customers. Yeah. John, Cloud, Cloud, it's sort of almost a veteran now, a couple of years of being out there. Yes. And I want to ask you guys. Yeah, so we basically have built a SaaS platform that delivers data science to the VMware admin, in a way that is highly simple and very rapid in terms of the time to value to figure out how analytics that are born out of the minds and work of very, very sophisticated data scientists can be put to work for problems that you currently face in an on-premise solution. And using SaaS is very key to delivering that value because it delivers simplicity and a rapid time to that analysis that you just don't see with traditional on-premise management applications where the overhead and complexity of those applications have prohibited getting there. The second part to this that's very important is as a SaaS application, we're collecting metadata from thousands of data centers today, and that creates a mega data set that effectively allows us to train algorithms that we use and harness the insights that they generate on your particular sets of problems. This is exactly what Google, LinkedIn, and all the very large SaaS organizations do with their operational management structures. And so we've mimicked that on the part of the VMware admin. Yeah, and that point is hugely important. I've been on sort of a soapbox for a couple of years saying that while a lot of this new technology is very cool and it has a lot of potential, the challenge is every single company has to build a learning curve as opposed to take advantage of the learning curve of others. And I think if I'm not mistaken, every single one of your companies over the last couple of years has won best of show here in one way or the other here. So that talent that we've always known from VMware is getting out there and you're now taking advantage, people take advantage of it in a different way. They can sort of scale it better. So Madura, talk about the first step that people have to sort of go through in terms of a lot of folks may be afraid of the public cloud because they think about when my data's going somewhere, I can't see it, I don't know where the box is, how do I audit it? But talk about when you're really talking about more of a control plane as a service, what's the thinking that helps them get over going? This makes more sense. Right, yeah, sure. So when we got started with the model that we were presenting to NFI's IT, our model is kind of unique because our control plane is hosted as a SaaS service, but it's still designed for management of private infrastructure. So users, storage, servers, networking, all that infrastructure lives on premise in their data center, but the management here, the control tier is outsourced to us. And when we started, we had a thesis, we thought that about 50% of the market would just completely want to stay away from us for security and control reasons. Or just general fear, you know, it's different. Right, exactly, general fear. But the other 50% would really love us because of the simplicity and out of the box experience that we offer. Right, the reality really has surprised us in some ways. We've been surprised by just how many percentage of customers have been open to our model, just as long as we build that trust between them and we're completely upfront and open about the security of the model and we kind of highlight various aspects of it. So the key things that are important from an IT administrator's perspective is first off, he wants to have complete visibility into where his data is going to live, where the metadata, if any, is going to live. What metadata are we tracking? How are we tracking it? How are we protecting it? Whether ours is a multi-tenancy model, whether we share end users data with other customers, et cetera. So what we do is we have these thorough conversations, we involve the InfoSec teams from the organization, the customer, and completely satisfy them and tell them that your data is probably even more secure with us than you maintaining it because this is our bread and butter. So we go out of our ways to maintain it in a secure manner. Right, well, and I think ultimately that's got to become a skill that every IT, Pat talks about OneCloud at Wikibon, we tend to maybe have some differing opinions on that. I mean, every organization today uses SaaS applications for something. It's HR, it's email, it's other things. You have to know how to interact with SaaS providers, whether it's for control plane or for anything. That's a skill you just got to have. So, Shiv, you're in a space that's being disrupted like crazy. I mean, SDN, hyperconverge, converge infrastructure, yours is challenging because you've got traditional silos. How do you help those traditional silos see, because now it becomes a little blurred, whether it's a hyperconverge, like talk about what the arch and tools do to give people a better visibility, maybe not what single pane of glass because that sometimes gets a buzzword, but how do you help them in that sense? So ultimately, it's what organizations are and what IT admins are, they are worried about, or they want to get visibility around is their workloads, right? Which is, infrastructure is just a means to the end and as infrastructure is changing, ultimately what will matter is that, is their application performing the way they want to perform? Are the virtual machines getting the right amount of security or not? So, the way we have been able to kind of get multiple teams on board is by showing them a picture, to show them a picture of how, given any virtual machine or a group of virtual machines which might be forming an application, how does compute, network security, all of it come together to kind of help achieve the end goal of that VM, right? So when they see this converged visibility, specifically when you are running converged infrastructure where everything is hidden inside a box and now you see components by components, how they are connected to each other, what's the relationship between them, what's the performance between them, it actually, that showing that picture helps a lot kind of conveying the value, the value prop and the message. So we pretty much our kind of demo does a job when you go into an enterprise and they can see the value right up front there. So it becomes about UI, it becomes about visualization, trying to help them see what that outcome looks like. Right, right, and when we show it, when you show it, you have to show the insights, the analytics as part of it, right? But you don't have to go deep into really, if you get into the weeds of, hey, I need to see the networking data and the compute and security and storage and all, then it goes into a different kind of conversation. Just show them the value, show them what problems will solve, it's much easier conversation. Yeah, John, you know, people tend to think about data science a lot of times with Hadoop and big data and these sort of magical, find a needle in a haystack thing. Help people understand like, how challenging it is to do data science and why, you know, when you're centralizing that and taking the learnings of it, that that power isn't just for data science for the marketing team, but how do you apply that for the IT team? Yeah, so there's some really primary use cases to talk about here in the application of data science and it really starts with what has been discussed here as the importance of the hybrid cloud. And if you start with just the basic decision of the formation of your hybrid cloud, you face a really fundamental decision on how to partition that. What workloads should I actually move? What should I order up from that hybrid cloud? And today, there's really no quantitative way to rapidly and easily understand which workloads make the best sense. And then you face this Goldilocks problem that you're pretty severely punished by if you don't get it right. When you order up an instance from a cloud provider to form your hybrid cloud and you don't get exactly the right size of that instance, you pay for that and you get a monthly bill that really calls out your mistake. So at Cloud Physics, the nature of how we both analyze your environment with what you have locally and using cohort analysis, which is typically used in a lot of data science applications to try to find similar types of configurations that have achieved just the right amount of resource application to make your workload run is an example of one of the first decisions you face if you're going to step into a hybrid cloud. So using that quantitatively to figure out, oh, these five workloads have this type of performance characteristic and this current configuration is such. You can use that output and that analysis that we generate for you to then figure out what exactly you need to order from your cloud provider as you form your public cloud. And that's, it's amazingly powerful how much better in real time we can get, you know? And you got to sort of bridge that gap. You might be buying resources by the hour, but you're getting a bill by the month and you've got to know somewhere in between. Like you said, you may end up getting a much bigger bill than you expected or something along those lines. So I'm going to ask each of you this question. So each one of your customers at some point is going to say, I'm willing to make that leap. All right, I'm willing to let you help me augment my IT, sort of get to that business outcome faster. What is, for the customers that you've got today, what does their world now look like? What do they tell you that, okay, this is, we're doing this differently, we're doing that differently, this is a different outcome, I'll let you start. Yeah, no, definitely. So we see it very clearly with our customers, right? Because our customers tend to be those folks, those IT administrators, et cetera, who have had prior experience with OpenStack. And OpenStack is an open source project and it's very powerful in terms of what it lets you do, lets you transform your infrastructure into an Amazon-like cloud. But it also comes with the difficulty, which is the pain point in setting up an open source project, right? It's extremely complicated, exactly. We famously say that it's a very powerful tool with the manual missing, right? So you really have to figure things out. So our customers tend to be those who've tried it on themselves for some time and then the projects get stalled, the people who set up the projects leave the company, you know, OpenStack expertise is one of the most difficult expertise to hire today. So when we come on board, what we promise to them is that we are the extension of your ops team and we are taking care of your OpenStack cloud. So we're de-risking your OpenStack strategy. And they very clearly see that. So it's not just about the initial installation of OpenStack, right? On an ongoing basis, we provide 24-7 monitoring and management and learning and they absolutely see that big difference. And that's a big deal. It's hard to find a lot of these skills where there's data science, you know, SDN and hyper-converged OpenStack, being able to aggregate that's a big deal. What are you seeing from the infrastructure space around this? Right, so one of the key things that we see is that we still see silos. And we see that there has been like 20, again, 20, 30 years of knowledge, certifications, training which has gone in each of these domains, networking, security, right? So, and these customers clearly see the value of software-defined networks of micro segmentation because it gives them more agility, better security, right? And companies like VMware and Cisco are doing the job for us on educating them on the benefits, right? So they buy into that value prop. Now the challenge internally is that, one, how do they bridge this gap between the different silos? Because if you want to adopt converged infrastructure or hyper-converged infrastructure, if you want to adopt software-defined network security, you have to have multiple teams work. You have to have your networking and virtualization teams kind of in the same room or your security team on board on to how security will be done, right? And I got to imagine a lot of it's just, how much can you get my ROI faster? Can I get that TCO faster? Exactly, and not necessarily taking those teams through the same amount of training and certifications. Otherwise you'll be parallelly creating the CCIE and CCN equivalence in the VMware world and in the software-defined storage world and all that, right? So the whole goal of ARKIN is that, how do you map SDN, software-defined security to existing constructs? How do you enable enterprises with existing teams, existing skill sets to monitor that, operate that? And the way we do that is by providing converged visibility and wrapping it up in a simple Google-like search interface. So what you do there is that, if you want to get any kind of visibility, you can use day-to-day data center lingo that in order to kind of formulate the kind of queries that you are and to get the information you are looking for. You don't necessarily have to know that it's a distributed firewall. A firewall is a firewall, a rule is a rule, a network is a network. So if you want to say, okay, what's going on between two virtual machines, you can just say, okay, show me what's going on between two VMs. We'll tell you, hey, this is the problem, right? Without going, making you to go deep into a NSX, VMware NSX specific pin dashboard or a UCS manager or a Nutanix specific dashboard, right? So we kind of bring it together in a very abstract, simple way and make it consumption also very easier for this. Everyone knows. Yeah, sorry. John, how much smarter do I get with cloud physics? What's an immediate way I can tell that that data science is helping me? That's, so I think the two emotions that we keep finding with our customers around that are surprise. And anytime you apply data science to any industry, you end up with surprises because you start to see things that you haven't seen before. And there's also an element of pride in that as you start to use cloud physics, you develop a baseline effectively that can be tracked over time to understand improvements that you're making quantitatively. That data can then be taken to your manager and you can actually show quantitatively that I took our baseline configuration and its behavior and look how much more improvement I made off of that fundamental configuration. So being able to talk with data just inherently improves the communication and the rationale for why you're doing certain things. You look smarter. You actually are smarter. I think because you now have data at your command that you simply haven't had in other products. And I think that's really the power that we're finding off of our platform. And it goes direct to the end user in what they do in their daily life. Right, right. So, we're going to wrap this up a little bit. The one thing we really didn't get into was the differing economics of SaaS. And one of the things that I've seen way too many times that I've gone into customer shops, you go into one of the offices and there's boxes and boxes of management software, whether it's network management or cloud management software. One of the brilliant things about the whole SaaS model is it's no longer ELAs. It's I can get a trial, I can get started immediately. I'm going to pay for what I use and you're going to get immediate feedback. That's got to be critical to sort of all your businesses. It let people see what you're doing. We should dive into that more into another session. I think we're going to run out of time here. Real last question, best way for people to engage with you guys. Just messages at platforms 9.6 on Twitter or write us info at platform 9.com. Likewise, we have a booth here. So, encourage everyone to visit our booths, see a demo of our product. We are on Twitter, it's at Arc in Net. And our website has all the information. Excellent. Same form factor you get with any SaaS model. You just go to the site and you download our V app and away you go. Okay. Or work with one of our partners who actually may be engaged with already. Okay, excellent. Each one of you, I want to thank you for being on the panel. This has been very interesting. Great stuff coming out of the VMware community. Great stuff coming out of sort of the V Mafia. Tons of brain trusts coming out of there. Lots of people making really cool things and really changing the way that we're going to engage around this. We see so much cool technology come out. We've got to be able to use it faster. People want to operate like the big cloud companies. You guys are giving them an opportunity to do that, take advantage of the skills. So with that, I'm going to wrap up. Folks, for everything that you've seen here, you can always look at the videos on siliconangle.tv. You can find all of our research on wikibon.com. With that, we're going to wrap up. Thanks again for watching from VMworld 2015. Continue watching. We'll have more all day long.