 From Austin, Texas, it's theCUBE, covering OpenStack Summit 2016, brought to you by the OpenStack Foundation and headline sponsors Red Hat and Cisco. Now here are your hosts, Stu Miniman and Brian Graceley. Welcome back to the Austin Convention Center. This is OpenStack 2016 in Austin, of course. This is theCUBE, it's getting towards the end of our third day of three here at the show. Happened to have on the program a friend from the community. Somebody is actually hard to believe his first time on the program. Tom Hollingsworth, who, in addition to being the networking nerd, is also the Dread Pirate Fosket, which means that he's part of the Tech Field Day and TFDX is here for the first time at OpenStack Summit. Tom, welcome to the program, and could you please give just a brief introduction to our audience about who you are and what brought you to this interesting role that you've got? Sure, so I'm a networking blogger by trade. I used to do a lot of work with schools and things like that and kind of started getting interested in writing down some of the stuff, so I didn't forget it. And then, you know, that kind of grew to the point where I just wrote quite a bit about technology changes, you know, those kinds of things. And then got involved with Tech Field Day as a delegate and eventually got to the point where I loved it so much that I decided to start doing it for my full-time job. Yeah, wait, wait, to tech blogging, that was something from like the last decade, I thought, right? I know, right? Because I keep hearing the blog is dead. But, you know, between blogging and podcasts and all the stuff, the videos we have going on now, you know, there's a lot of ways that we can get information out of the community and we're always trying to find new and interesting ways to tell people about, you know, what's going on in the community and technology in general. All right, so for those that don't know, could you introduce Tech Field Day and what brought it here to the OpenStack? Well, so Tech Field Day is a really interesting concept that we have with the community where we go out and we kind of, we foster the people that are wanting to learn more about technology and to share their vision with the rest of the community. And we get them in a room with companies, you know, a lot of the companies that are here presenting at OpenStack Summit are also companies that have presented at Tech Field Day. And we just let them get as deep into the technical conversation as you can imagine. I don't think we've ever had anyone come out and go, you know, that was probably too technical, we shouldn't have gone that far into it. But it's great because there's so many things that we've seen from companies, you know, storage and networking and compute resources. And a lot of that just gets referenced other places. I was looking up something on the OpenStack Wikipedia page yesterday and Tech Field Day has a link on there from one of our presentations. Awesome, so we want to dig into some of the networking stuff. I'm real happy this year we're not discussing kind of the stability of Neutron. We spent the last couple of years, you know, it was there and who's working to fix it and who's fault was it that it wasn't working right. But there's been a lot of networking discussions. Of course, those wonderful acronyms, kind of SDN and NFV, NFV being, you know, a big, big use case talked about in the keynotes. As a networking person, Tom, what's your experience been this week? What's been interesting? What surprised you? What hasn't surprised you? Well, like you said, I'm so happy that we've moved away from fixing the broken things in the project to using it as a platform to launch new things. It means we fixed it, right? We hope. Kind of. I think what you're gonna find moving forward is that because really, I don't necessarily know that it was much that Neutron was broken as much as networking didn't know how to take advantage of what Neutron was offering. And so now that we're starting to have bigger conversations around programmability and APIs and automation and orchestration and all those SDN-y things, now what you're starting to see is people are trying to figure out how to drive that into projects, like OpenStack and all the other things that come out of it. So what I think that means is that we're gonna start getting away from the mentality for networking we have now where it's a bunch of guys that are just waiting for a VLAN and it needs to be programmed and those folks are gonna start doing better things with the underlay technology, like storage and compute resources. Those kind of disappeared into this web of management consoles and stuff like that and the hope really is that networking will eventually do that. Yeah, so one of the keys and when you're talking about networking, tons and tons of acronyms and so forth, break down a little bit for us, for anybody who doesn't live in the networking world, the difference between networking on OpenStack and then we heard things like NFV, we saw some folks from the Open Daylight Foundation here, what do those different things do or what's their role? Yeah, so networking really for the last 10 years has been very stagnant. Things just got faster but they didn't get easier to use. And so once we started figuring out that we needed orchestration layers and we needed things to work on, that's when people got really heavy into development around things like NFV, which is network function virtualization where we take something that used to be an appliance for packet sniffing or network monitoring or even load balancers and firewalls. And instead of having to run that on custom hardware or even generic hardware, we turned it into a software layer. And so then that lives in the cloud but it doesn't mean that we have to re-architect the entire system in order to put something in because I'm sure you've dealt with that before where it's like okay we got to put these load balancers at this spot because if we don't do it that way it won't work right. With NFV it's all like I can magically move things around as I need to. And that really works because we're starting to figure out how to make software do things better. And so you mentioned Open Daylight, the Open Daylight Foundation is a wonderful group because what they're doing is they're trying to create a controller architecture for network programming where instead of having to go touch 15 switches to update rules or things like that I have one central control unit that will do all of that for me so that I don't have to spend my time replicating my changes over and over again because we all know that the likelihood of making a typo as you go forward increases substantially from like switch 12 to switch 13. So it's like yeah, we got to get away from that mentality really. And that really highlights why they're talking about McKeenotes is saying like OpenStack has got to be collaborative. They've got to collaborate with the guys doing network programmability and virtualized functions and all those things. Yeah, and really I think that that's super important especially for the networking community because that's not something we're used to. And so if we don't get an opportunity to experience that and to see it become part of a larger organization then no one would really understand the importance of it. I mean, one of the favorite things I love to tell people is I want you to go into your AWS console and I want you to create a VLAN. And most of the networking guys are like, oh, I'm going to figure this out. And the computer guys are like, that's not an option for me. I'm like, that shouldn't be an option. If we got to that point, we broke something somewhere along the way that didn't need to be broken. Yeah, so does that mean all the networking people are replaced by robots, Tom? Well, I'm working on a project on the side. And now the funny thing is people love to say, oh, my job's going away. My job's going away. And while I will admit that the world is still going to need some ditch diggers to rack and stack equipment, ask those server guys that used to manage 15 or 16 Windows boxes what happened to their jobs. They're still working somewhere. They're just working on something different. And I think the smart people, the people that come to things like OpenStack Summit are going to learn how to retrain and they're going to learn a new way of doing things, a better way of doing things, and they're going to be happier for it. Yeah, as you talk about Tech Field Day, you've got vendors and companies coming in front of your panel. How is it different or the same now that you've got a mix maybe of vendors who did proprietary hardware and software versus those who are mixing in open source software or open sort of proprietary, not proprietary, white box type of hardware and stuff. How does that change? It changes the conversation because in the past, it was what's the secret sauce inside of this magical little box that makes the packets go where they're supposed to go. And we spent a lot of time focusing on that. But now that we're starting to see people adopt open concepts, open ability, the conversation shifts away from the magic stuff and more to how are you making magic with it? So now it's you go, with Juniper doing OpenContrail, it's like, oh, well we understand how that works. What are you doing with it? So that leads to disaggregation, that leads to orchestration platforms and that kind of thing. So now we're understanding more of a basic level so we can cut past the speeds and feeds on a box basically. So one of the big discussions here has been that NFV use case. I'm curious, just from your viewpoint, the general applicability, kind of the size of that opportunity and how that fits into the overall discussion of networking in the open stack space. Well I think there's some real opportunity for creating distributed systems through NFV. Because for years, we've worried about make the box as big as you can make it because we need to be able to throw 10 million packets per second at it. But I talked to a guy earlier and what he was talking about was using containerized firewalls as a distributed IDS. So now I can put IDS sensors everywhere I need them but instead of each one of them having to have the intelligence to process things, they can feed all that information back somewhere and analytics can say, well you're starting to see pieces over here. That's not possible with hardware today. NFV is huge because it's going to shift the way that we think about how we can deploy devices because when it's all software, the 15th IDS doesn't cost any more than the fifth because it's all just a little extra. And I'm sure you're starting to see this when you talk to companies that they have an app store model for this kind of thing where, click, click and hey, I have a whole bunch of new devices in my network. Yeah, now we heard the NFV discussion mostly around telco, so AT&T and Verizon. Is it applicable to the non-telco world as well? The enterprise government agencies? That's like that. Yeah, it's absolutely applicable. The service provider argument is fairly easy because that's just additional service revenue that they can provide for customers that don't understand this kind of thing. So now it's just like ordering from McDonald's. It's like I'll have one of those and one of those and one of those and you're done. But when you have the ability to do this in your own enterprise, when you're like we're no longer constrained by needing to spend CapEx dollars on whatever this is, that gives you flexibility to do so many more things inside of your enterprise. And when you look at the shift of how enterprises are starting to look like service providers to their users, then it gives them a whole bunch of additional capabilities and services that they can then offer and provide value for IT instead of making it a cost center like it has been for so many years. I'm curious, Tom, when you get to talk to the practitioners and the users out there, how much does open in the networking space matter to them? I know you've gone to the open networking user groups and open networking summit here at OpenStack, open, open, open, but where does that fit on the priorities of the users? It really fits with a lot of the priorities, especially for the folks who are doing like the advanced operator stuff. Guys like Jason Edelman and Matt Oswald, couple of great tech field day delegates that have taken what they've learned and expanded it to include so many more things, DevOps for networking or the orchestration platforms that they're working on to build things better. Those don't work unless there's a way to open up what you're doing. I mean, for years, IT has been like, let's make it and then let's not tell anybody about it because if they copy it, it's not going to work right, but no, we have to share everything because if we don't, then you've got a one-off box that you wrote code for and you're either going to have to support it for the rest of your natural life or you're going to throw it out in five years and you're going to be mad you had to buy a new one. So as an old CCIE, if somebody came to me and said, hey, you need to be a programmer, stop being a CLI jock, you got to be a programmer. If you didn't go through CS or you weren't a programmer, you sort of blank out, you go, okay, where do I start? Have you seen any good examples or just kind of ways that people have gotten started where you start saying, hey, you're still the network person, but you've got to start automating what you do. How do you deal with that? Yeah, there's a lot of great resources out there. They're actually, they're starting to write books, Python for networking programmers because we know what we need to do as a networking professional. We know what information we're trying to look at or what commands we're trying to send. So if we can figure out how to do that in a slightly different syntax against an API, then that's half the battle right there. So there's a lot of great resources, you search them out and the wonderful thing is is that because of the community that we've got, a lot of those resources are available for nothing because we want to share back, right? Builds the community faster. Oh yeah, absolutely. So Tom, like us, you get to a lot of events like this. Somebody I was with this week made the comment that kind of resonated with me is that this is an industry show, run, not by a vendor, but it's a community show and it's not only growing, but it's thriving. We saw so many of the shows. So what makes this show different from some of the other ones out there and you know? It's funny, something you mentioned earlier. It took us a long time to get here. Not necessarily because we didn't like OpenStack or something like that. We really didn't understand it. And so for us, it was when everybody I talked to said, hey, are you going to be at OpenStack next week? Should I be? And so this year I was like, I think we should be. The users, I mean, so many people walking around not wearing corporate logo polos or things like that, they're invested in the community and you can tell that they want to share, they want to give back. There's a lot of great community stuff. V Brownbag has got a room that they've gotten booked solid for two and a half days of community presentations. There's a real feeling that people hear matter. The things that they do matter. You're not just a customer or a dollar amount to somebody. It's like, I wrote the code that built that object right there. That's important. Yeah, there's definitely everything and there's a lot of the hashtag we are OpenStack as part of it. What's interesting is a couple of years ago, it was all people that were helping build the code. And one of the comments I made is yes, many of the people that are helping build the code are also the people that are using that code now. What do you see the blurring of the lines between users and creators of technology? Well, for engineers, it's very easy for us to get focused on the tech. I have to build the perfect protocol or I have to make this ultimate cool box that does things. And we forget how to make applications of that technology. And I think what we're starting to see from the community is that as more folks are getting involved, maybe they're not the best coder in the world. But boy, I figured out something cool that I can do with this tech. And that's what's driving enterprise adoption of it and what's increasing the user base. So now as they get invested in it, that's when they start picking things up. More to your point, how do I learn how to contribute code? How do I do things the right way so that my contributions can live on and make somebody else's life easier? So I think back to this whole journey to cloud. Networking has traditionally been one of the things holding us back. Where are we? What's your commentary on that? It's so tough because with compute and storage, they had similar traits. So when the judgment came that we're going to do more software stuff and we're going to go to commodity hardware, it kind of hit them both at the same time. But networking resisted that because we have custom Asics and there's just certain things you can't do with general purpose CPUs because they're too slow. But as we're starting to use software as a driver to create change, now networking is starting to fall into that same umbrella so that it's a utility resource. I mean, I don't sit at home at night and wonder how the electricity gets to my house. I know what I can do with it when it gets there. And that's what networking needs to be for the organization. It needs to stop being a hobby kit that breaks all the time and it just needs to be a utility resource that's available for other people to use. All right, so Tom, I want to give you the last word. Anything you want to share, kind of coolest thing you saw this week, something interesting, just fun times in Austin. Well, I mean, I love Austin. It's got a vibrant community and it's so amazing to have people here and then to have all these folks just walking up and down Sixth Street talking about cloud deployments and things like that was kind of weird when you consider, you know, keep Austin weird. And I know I can't think of a weirder way to do that than to have OpenStack here. But I'm excited because I feel like this community has the power to change a lot of what's wrong in IT. And the fact that it's grown so huge so quickly means that other people are starting to recognize it too. And I think at the end of the day, what you're going to see is a lot more people getting on board, a lot more people contributing to the project. And at the end of the day, hopefully some heads will turn in the important parts of the organization to say this is something we need to take a look at. All right, well, Tom Hollingsworth, happy to be able to call you a CUBE alumni now. We'll be back with lots more coverage here from OpenStack 2016 in Austin. This is theCUBE. It's always fun to come back to theCUBE because, you know, the discussion is always interesting and relevant. It's not scripted, which I think makes it real. And I think it's really good service for the viewers who actually get that insight and the ability to understand the conversation and dialogue. And it's actually for the folks who are invited, as I've been invited back a few times. It's actually kind of exciting because you never know where it's going to go. Hi, I'm Scott Nal, CTO of Hortonworks, and you're watching theCUBE. The industry, especially the big data industry, and also they, this is Wei Wang from Hortonworks. You are watching the CUBE. Hi, this is Krista Vanny from Data Robot. My name is Nenal Paxada. Hi, this is Wei Wang from Hortonworks. Hi, my name is Amitwalia. I'm the executive vice president and chief product office at Electing for Manica. Hi, my name is Rithika Gunnar. I'm the vice president of Data and Analytics at IBM. And you're watching theCUBE. You're watching theCUBE. You're watching theCUBE. You are watching theCUBE. Summit 2016, brought to you by the OpenStack Foundation and headline sponsors Red Hat and Cisco. Now here are your hosts, Stu Miniman and John Walz.