 And we're back. This is Stu Miniman with wikibond.org with SiliconANGLE TVs, live continuing coverage of all the best tech action here from the Open Networking Summit Center of the Networking Transformation. And joining me for this segment are Brent Salisbury and Madhu Venekapal, both from Red Hat. Of course, we're talking a lot about open source, so it's only fitting that we get Reddit in here. Guys, thank you for joining us on this segment. Sure, thank you for having us. Yeah, absolutely, man. Great, so Open Networking Summit is about a lot of different projects that are going on in open source community in general. Both of you are relatively new to Red Hat, but you're both very active in the Open Daylight project, so you're going to go into it a little bit. But Madhu, maybe you can tell me a little bit about your background with Open Daylight and what brought you to kind of your current role with Red Hat. Yeah, sure. I've been with Cisco for a very long time. And around 2010 timeframe, I was really bored with Networking as a whole. It was stale, it was stagnant. I was trying to do the same thing again and again on various versions of the Silicon. So then I decided to actually quit Networking as a whole and go into the app space. But I got lucky in 2011 to attend the first O&S right here. Actually, it was in Stanford, actually. There, I attended all these sessions by Martin Casado and Scott Schwenker. It totally inspired me. And after that, we started working on a control project back at Cisco that actually helped with Open Daylight into the contribution. So I've been working on Open Daylight, for Open Daylight, base code from 2011 onwards. And last year, I had to join Red Hat because I wanted to be in the open source way of things. So did you guys bring your little coin, the code, from Open Daylight? We had Dave Mayron earlier talking about, of course, for open source. It's not just being part of the community, but coding is part of it. So both of you committed a lot of code? Oh, yes. And we have the coin. The coin is one of the best things happened about that in my career, yeah. Yeah, all right. So Brent, one of the reasons I wanted to bring you on is to give a little bit of that end user viewpoint. So while you were with Red Hat now, you were at University of Kentucky before. Can you tell us why you got involved in Open Daylight, that project, and why did the University care at all? Were you just doing it on your own? How'd that fit? Yeah, I mean, research and education has a little bit of, they've got plenty of interest in it. A lot of it's from the research side. So I was on architecture operations. I saw the pains up front. Been doing this for 15 plus years. Nothing's really changed. So we can only go through the same problem over and over and over until it's either you just quit the industry or you're all together, go do something else. So it's kind of that time for transformation. Wow, so I'm hearing from both of you that networking has kind of stagnated. Do we feel that with the open networking movement that networking's reinvigorated and something that excites you and should excite people? Yeah, I don't know if we're just freaks of nature. Well, we are freaks of nature, but we're starting to see more and more people from traditional networking come over to, I think there's a little misconception on yes, code is obviously incredibly important to open source projects, but I'd say for every line of code that we push, there's probably eight hours behind it, banging your head against the wall, doing integration. Everything's so new, between the data path, the control, proper architectures to put in. There's so many places for traditional networking to, and it's the best thing you can ever do for your career. Yeah, so the open networking is the best thing that happened to networking, as in a long while, we all agree on that. As a developer from my early career, I feel that this is the first time I'm enjoying the coding, per se, because the architecture is completely different from what we used to do before. It is centralized, but not completely centralized, because we have distributed architecture, clustering and everything. So we feel like programmers now, before we feel like network engineers morphed as programmers, but now we are hardcore programmers doing the real things like real programmers do. So it's fun. Yeah, actually, Brian, I'm sure you've seen, there's been some articles out there, do networking people need to become coders? And there's a lot of people that got into networking because they didn't want to code. What's your take on that? What are you hearing your peers in the marketplace? Right, so to manage a virtual x86 environment, do you have to be a programmer? No. To manage a wireless environment, do you have to be a programmer? No. So these are all things that have some form of fashion that, more than anything, we're decoupling things. So we've got a monolithic box that's black and there's x86 on it. So it's time to kind of break that apart. Yeah, Madu, I'm wondering if you can tell us, how much networking is there inside a Red Hat? I mean, obviously Red Hat's heavily involved in a lot of open source projects. Where do you guys sit in the organization? Why is it strategic? To what, you know, brought you to Red Hat, what Red Hat does? Yeah, good question. So Red Hat has been investing on OpenStack a lot and in the networkization space, we need a solution, a good solution on that. So that's when it's a good combination for Red Hat to be part of OpenDLite, to promote open source and also do something really good on the open networking space. And the reason I personally joined Red Hat is that I was looking for an avenue to be a part of the DNA structure which can do innovation through openness. And I found that Red Hat is the perfect spot for me to be a part of. So, yeah, I mean, Brent, would you recommend kind of career path is, you know, we need more people coding? My understanding is OpenDLite's looking for people, what's the pitch for people looking at this? Yeah, I mean, you know, the beauty is your resume is whatever you've done in the past. So you're either showing some, you know, either repository or code or you're jumping on IRC, working in open source projects. You've got OpenStack, you've got OpenVSwitch, you've got OpenDLite, you've got, you know, OpenFlow isn't necessarily something you go work on, but you can always do that at your shop. So, absolutely, if you go and grind for six months, you can go basically get your dream job wherever you want. It's not going to last forever. But, you know, I mean, the funny thing is everything we're doing, if you look back at history, we've already kind of done that, right? So we come out with, oh, hey, having a decoupled operating system, that's incredibly innovative. How many server guys do you, you know, think about where you're going, hey, you're going to have to buy Windows and you have to buy this piece of hardware to buy it on. It's kind of bizarre. All right, so let's unpack ODL, OpenDLite for a little bit. Madhu, you know, you've been involved for a long time. You know, some people have said, you know, gosh, you know, this program's been going on for a while. You know, why isn't it in production in, you know, all the data centers around the world? Of course, we know code takes time, maturity takes a while, ecosystem takes a while to build. You know, where are we with, you know, OpenDLite and, you know, what, can you give us the update on OpenDLite? Oh, absolutely. I mean, OpenDLite is part of, for a while, only from a perspective of, the contribution came from a big incumbent from the past. But it was looked at from a different perspective before. Thank goodness, the OpenDLite source code originally was in the proper architecture, so we can fit in various use cases on that. So original use cases that was addressed on the Cisco site was not, especially on the data center side, it was on a very small use case we are addressing on that. And when OpenDLite started last year, it was around, it was been 10, 11 months now, then it has actually expanded after that. So as you know, the past 11 months has been the time when use cases are actually coming and contribution has been, you can start seeing that. So hydrogen is the first step towards that. I would never say hydrogen is the ultimate release of OpenDLite. It is not, it has been many projects that has been, not even core projects, they are part of it now. They are all, you know, incubation projects and bootstrap projects. So the quality is not there yet. There's a lot of features, a lot of the functionalities. Folks can pick it up and try it out on the lab and you know, get the feel of it. But there'll be next releases, helium in six months. You'll start seeing real things going on. I mean, you'll see good quality code coming in, real use cases you'll start seeing pouring inside. So is it going to be on like a six month release cycle? Is that what we're looking for? That's the plan as of now. TSE is open in OpenDLite and everything is open in OpenDLite. It's a pretty cool thing where that's like the release cadence we have so far is going to be six months. Since we have first release done, the next release is going to dictate how much it's going to be. So it's going to be six months release cycle. That's what we hope for. Yeah, so Brent, you know, you've done a lot of writing as to how people get up and started on this. You know, what's the barrier to entry for somebody to, you know, get started? How easy is it? You know, where would you recommend somebody start? Yeah, absolutely. You know, a lot of our focus is at the soft edge. So anything that's virtualizable today. I'm sorry, explain what the soft edge is. Oh, sure, sure. Right, so the servers, we've got hypervisors. We've, you know, we're commoditized. That's been commoditized for a while. Red Hat SKVM fits perfectly. You know, I love Red Hat because what they do is take proprietary concepts and create platforms out of them that other people can come in and innovate on. So just getting to open source in general, getting to know how to push, you know, scripts, code, anything to a repository, learning any kind of CI tools, you know, just learning a v-switch, incredibly important. Python, Bash, Linux, all of that's great to start out with and then, you know, you can move into lower level languages as you keep going. All right, so I wonder if we could shift a little to, you know, OpenStack. You know, you got, are you a little bit involved in that? Oh yeah. I think you are, yeah. No, I mean, that's been our pretty much obsession for the past few months is getting integration in. We're still developing out features, but, you know, the funny thing is that, you know, there's not a whole lot of alternatives when it comes to a lot of these things, right? So, I think the industry is pretty smart to eventually kind of consolidate on a project and let's run with that and kind of focus. I think it probably reduces the risk to a lot of the income. So that's the open source option is what you're saying, because obviously there are a number of, you know, cloud options out there. You want to tell us, I mean, you've been kind of hot on the OBSDB, you know, what's going on there? Yeah, so we've implemented, so Nassir, VMware, Ben Faf, in particular, working for marketing setup together at OBSDB. He's pretty incredible. So we're just implementing that protocol. That's a management plan protocol. So, you know, I know I've talked about OpenFlow a lot. Madhu's been working on that project for, you know, a couple of years. So that's just one piece, though. So it's a whole lot easier if you can start leveraging kind of multiple angles of the fabric. And so you don't necessarily get everything you want out of just something like OpenFlow. So now you can kind of cross reference that. It becomes particularly important when we're coding. So you've got basically OpenStack coming in from one direction. You've got OBSDB coming in from another and the controller's coming down, giving you more information. On top of that, you've got a northbound API that's feeding you information. So a lot of air checking and a lot of correlation there. Yeah, so I had Dave Meyer on earlier and he, you know, was very clear about the openness of the project and how the APIs would necessarily, you know, would need it to stay open. But, you know, I think that's one of the concerns people have is, you know, how open this'll be. You know, Madhu, can you give your perspective on, you know, what you've seen both, you know, you're on the Cisco side, now you're on the Red Hat side, you know, heavily involved, you know. Is this just a control play from the vendor community or, you know, is it truly going to be an open project? I sure hope only not because, if you look at the OpenDialite as a community, right, yes, it is made of a lot of engineers from various vendors, but end of the day, we're all developers working on a common cause. And if you log into the chat or if you look at the emails, look at the TSC calls, it's very clear that it's all open, it's all about open. It is nothing though both the community and everything's about the community really. And the vendors actually are doing a great favor by giving resources and great engineers to the project. And everybody's trying to find their place in SDN and everybody's trying really hard to be part of OpenDialite and doing great things. And Red Hat, as you know, it's a, they can shepherd open process projects really well because it's a bread and butter. And we being part of Red Hat is good for us, unfortunately, because we are really seeing what is an open-source way of doing things. Also, we can coach people who are from other vendors to see exactly what is open source and how to play in the open source as a community. So it's a win-win. So OpenDialite is one of the most open projects you will see. And whatever you hear, it's exactly what is being done inside as well. So TSC calls, TWS, everything's open. And me as a developer, I'm completely into the upstream. There's no downstream work that I personally work on. Same with brand. So it's been a fun ride for you. Great. So Madhu, I'm curious, are you guys involved once the first release is done? How does that go into the services and the products, both Red Hat and engagement with customers? Is that something you guys are involved in, or do you help with the training of other folks inside your organization? Oh, good point. I mean, first release, as I explained, the first release is not going to be the production release. First release has been a great place for us to even talk to the customers, talk to anybody that we partner with. At this point in time, we, myself and Brent at least, we are upstream guys at this point, and we are helping our management to explain what is SDN because SDN is new to all of us, including Red Hat. So we tend to explain where Red Hat can play a role and how we can really help the community and the customers to break this ice on SDN, coach them on what is SDN as and how we can make a difference. So from a product standpoint, the CTO office will actually work with their product, isn't it, Red Hat, to see what we can do with Open Daylight, maybe down the line, once you get a good solid base, yeah. Great. So, you know, you've been meeting with people at the show, you've been going around, what have you seen so far? What's your take, Brent? Well, we've been probably nonstop doing demos on stage and presentations, so we're both probably pretty braindead at this point. We met with Martin this morning, he's just incredibly inspirational about that, so that was pretty awesome. I think we're still defining exactly what it is we're trying to, you know, I mean, the clear, the killer, so if we're looking for the killer application, it's pretty obvious that it's gonna be network virtualization at the gate. That's not a new idea, that's something VMware has paid plenty of money to get. That killer application of that application is pretty apparently gonna be OpenStack, so. And I think the long tail of OpenStack is, as we have, as NFV kind of crops up, which is becoming network function virtualization, where you take x86 compute off the shelf, you basically drive traffic at the edge of networks through it to do traditional black box services, low bouncing firewalling, all the insane carrier acronyms. So that I think is gonna just incorporate OpenStack, because once we nail down all these services, now we can just distribute it out to Telcos, Pops, wherever it needs to be, and there's real capex and not-bex savings, because to do SDN, it requires storage and compute orchestration also, it's kind of the key there. Network-centric approaches, just our network-centric approaches, we've kind of had that. We don't need to reinvent the wheel completely, but I do take-aways from the show. The show? Yeah. So, Qube, we have been watching Qube for many, many interviews, right? And Brent has been talking about you a lot, and whenever I watch the Qube interviews, I get inspired, especially whenever I see Martin, I sort of talk about that. Martin is my hero, right? I mean, it's obvious. And we've watched many interviews, and especially Martin's interviews, they're so cool, where he talks about his team a lot, right? And the way he's down to earth, and the way they have opened up the data space, and thanks for you guys for bringing them expressions to us. We can see them as individuals, not as the city-votes of the companies, right? So, it's inspiring for all of us. It's really inspiring for all of us, and thanks for you guys for bringing us. Oh, no, okay, but hey, we appreciate you guys coming on and support. We've got a lot of open-source coverage. I'm excited that the Qube, we're going to be in April, we're going to be at the Red Hat Summit in San Francisco at Moscone, which is the 10th anniversary show. In May, we're going to be at the OpenStack Summit. So, absolutely, we're doubling down on our coverage in this area. So, just guys, I want to give you kind of last closing thoughts as to meeting with customers here in their take on ODL and what's going on there. What do you think people should be looking to do in taking away from this event and what's going on recently? Madhu, start with you. Yeah, so one thing I wanted to shout out to the interior community is that open-source project like Open Daylight is not just for the programmers. It's for network governments, network architects, and so on and so forth. And the contributions can come from any place. We're not from the developers alone. It can come from network architects who can come and give their use cases to us. It can come from you who can promote what we are doing out there. It's really open to remove all the doubts from the customer's minds and so on and so forth. So, Open Daylight is one of the best projects out there and my wish is that everybody should come and participate at least on the TSE calls and TWS calls to know really how open it is and give a contribution. Go out otherwise. Right? Yeah, no, I mean it's, where else? You know, in the history of computing, of network computing, can you just say, I want to go work with the best of the best in the world. You know what I mean? It's worth it. And I'm a brother here, I'm a dude. I mean, it's pretty phenomenal. So, I just can't encourage enough for people in networking today to just, you know, at least step outside the box and evolve a little bit. You know, at the very least you're going to just, it's going to open up new worlds to you, whether it's jobs, whether it's intellectually. You know, there's opportunities here. It's just up to you to do it. Yeah. So, Brent, Madu, thank you so much for joining on this segment and love to get your viewpoint where Red Hat's sitting at in the community. It's about community, it's about code, and it's about transformation and networking. This is Stu Miniman with wikibon.org. The Cube will be right back with our next guest from Open Networking Summit 2014.