 So, thank you all for staying for the afternoon. We are closer to beer than we are breakfast, so we're making progress. Real quick, so again, thank you all for staying. I know it's tough to be here for the whole day. Are we having a good time? Has this been useful so far? Good, yeah, okay. Real quick, give Diane a round of applause. Putting this thing on is a huge feat. So, we've had a lot of really good technical panels so far. We've talked a lot about why people are building, what they're building, the thought process behind it, the engineering that went on behind it. This session is going to be a little bit different. We're going to do a panel session, but we're going to talk a little bit about, you know, all of you here, there's a reason you're here. There's a reason you're here because you're looking to do things to help your company make more money because you're going to develop a very cool new application. You're looking to, you're part of the sort of ecosystem that's doing technology. And to a certain extent, how many folks are staying for the rest of the week for KubeCon? Okay, so a good percentage of it. So, there's a part of what you're trying to get out of all this beyond the technology that is, do I want to be part of this community? Right? And this community being a subset or a super set of the Kube community. So, we want to talk a little about what that means in this session. And so, first I'm going to introduce everybody on the panel. In fact, why don't you guys go ahead and introduce you. You can still do a better job than I am, but go ahead and introduce you. We put this panel together. We wanted to have sort of end-user customers and sort of unique end-user customers that are helping to contribute. We have folks on the panel from some of the technology ecosystem partners and also people that are in sort of the SI world. So, you're seeing sort of the ecosystem, the supply chain, if you will, for doing this. So, Christophe, go ahead. So, hi everybody. So, my name is Christophe Atyas. I work with Amadeus. So, first I apologize for my French accent, but you will have to bear with it. So, Amadeus is an IT company providing, IT services for the travel industry, like airline booking systems, reservation systems. And we do, we are in the process of migrating all our applications to OpenShift, OpenShift running on OpenStack, and we are in partnership with Red Hat to work on OpenShift V3 very intensively. Yeah, cool, great. And I'm Jeff McCormick. I work for Currency Data. We're a postgres company. We primarily do support for postgres. What we also do, though, is containerize postgres and postgres technologies specifically. And we have those running on OpenShift, Kubernetes, and Docker. And anyway, that's what we do. I'm Chase Stewart with Arctic. We are more on the SI space. We're actually probably one of the newer sort of members of the Commons group. So, we're the newbies. Our focus typically is around consulting services with Red Hat products. So really what we wanna do is help take OpenShift and satellite and cloud forms and all the upstream components and help our customers sort of have a great experience with that, but at the same time, we've been surprising, surprised since we started that most of our customer bases ended up being sort of enterprise grade customers, large financial firms, things like that, which when you try to take their standards and take new technology and try to marry them together, there's a bit of glue that needs to be done. And so really our focus is helping them step ahead while making that sort of process smoother and sort of helping them develop some of the glue. I'm Rob LaLonde. I'm a general manager of Univan NavOps. Christophe apologized for his French accent. I'll apologize for my Canadian one. Hopefully you can understand me. But we have essentially a schedule and policy system for Kubernetes. So it's on top of OpenShift or any of the Kubernetes distros and provides sort of advanced workload management. Very cool. Gentlemen, thank you for being here. So one of the tricks is when you wake up in the morning and you look at the schedule and you realize that you have to go after Kelsey, there's a little bit of anxiety that takes on. And you wonder, how am I gonna keep people's attention, right? I was a little bit jealous that Alex was able to pull Kelsey into a session. So I thought, all right, I got a bunch of questions prepared, but how do I make them interesting? So I thought I'd try this approach. If this works, can I get the button to work? Or you can advance it if you want to. It's a bad approach. It's a bad approach. All right. So yeah, you wanna try it? We'll move on. All right, so let's start with this. Let's start with this, real simple. What's, you're all part of the OpenShift ecosystem in some way, right? Whether it's new to the community, whether it's been working on it for a long time. What's working? Anybody jump in, but what's, what do you find that's working so far? At the end of the day, you go, this is making my life easier, right? Sorry, just to keep it easy. Sorry, it's, as a newcomer to the sort of to the group, really, this is about connecting people with other people. And so she's been fantastic to both help connect us with other organizations that might be able to help us out as well as connecting them with us, where we can be of assistance as well, or asking us to participate in things like this, or lead events, or do demonstrations, and things like that. For me, the community is about educating, and she's helping us be smarter. Yeah, I definitely would agree. The people at Red Hat have really sought out to partner with us, and bring us into getting the latest information, the best sources of information within the, like for me, the engineering group, and having real direct access, and really the openness of that give and take process has been a good experience. I think that works very well. Celebrate on that, when you say direct access, what does that mean? Well, for me, dealing with like persistence issues for database containers has been a big deal, and being able to directly interact with like Kubernetes or OpenShift developer, who's working upstream on that particular issue, it's really been a lifesaver in a lot of cases in helping me figure out what I can do within the technical space. So, you know, them taking the time to make themselves available, answer questions, has really been a great experience. Yeah. So, yeah, for us, it's a bit the same. I mean, the collaboration, we have some people that are embedded in the OpenShift development teams and the Kubernetes community as well, and being able to work with them on a daily basis to make the product progress improve and so on. It has been really great. It's working very well for us. Yeah, no, you guys are a little bit unique for folks that don't know. You're essentially the backbone of the travel industry for Europe and a lot of the world. I read recently, your load is up about 6,000%, about 6,000 times what it used to be in terms of people searching and so forth. Something like that, rough numbers. Talk a little bit about you are not just consuming the software as if we used to with vendor, sort of the vendor customer relationship. You're now writing code, being part of that community. We've got a chart that we use all the time that Amadeus is actually one of the, you know, the bar graph. Talk about how that happened. How do you go from probably consuming proprietary render code to actually contributing back to communities? Well, it's all about the partnership with you guys. I mean, it has really started with us having a problem statement, which was, you know, making our applications more resilient and benefit from what, you know, the open source community was providing, so Kubernetes OpenShift. We started out with, you know, working with you guys and we found a very good model by having people being embedded from Red Hat with us and us with Red Hat and that works very well. I mean, it's how it has been set up for us. There's a lot of communities. Kelsey just got done talking about it. You guys have a lot of choices. There's a lot of open communities. There's a lot of opportunities for you to say, how do we want to go to market? Who do we want to work with? Why are you here today? Why would you choose this community, the Kubernetes community, the OpenShift community versus other communities? Because your time's valuable. Well, for us, our customers really moved us towards OpenShift in the first place three years ago and our customers really, we saw adopting this technology. So for us as a company, it was pretty easy to say, hey, that's where our customers are and we felt like we saw a really good collaboration between Google and Red Hat on Kubernetes and we saw that's really, we think, gonna bear a lot of good fruit and that's where we wanna place our technology kind of bet. And I think two and a half years down the road now it's really proven out to be a good bet and the way that community worked all along gave us more of a comfort level that, hey, this is gonna be successful and we wanna be part of this and we think our customers are gonna be successful on this platform. So for us, it was about a two and a half year. Let's watch this happen and it did happen in the way we wanted it to turn out, I think. Now clarify something. You guys are fundamentally in the persistent data, persistent stateful application space. You know, we hear a lot about 12 factor applications about all these sort of things. Like what, you said your customers are driving you towards this. What about their applications and this sort of hybrid platform thing makes sense for them? Customers are in at least some of the customers on my mind right now are kind of, they saw a lot of benefit in deploying web applications in a container environment and that's really easily demonstrated but a lot of these customers have very large database deployments, not so much in maybe the size of the actual data but some of these customers might have 50 test environments, 50 QA environments and they may need to spin these up very rapidly. So for us, two and a half years ago, we said, wouldn't it be great to have that same kind of deployment agility that you get with deploying a web app in containers. If we could get at least halfway there with deploying something like a database in that same agile manner, we knew our customers would really benefit from that and like I say, early on it was kind of harder to see databases fitting well in that model and there were some growing pains for sure but I'm happy to see a lot of work in dynamic provisioning and Petsets and stuff like that. That really helps us deliver that same kind of agility for the database side as you would see with web apps. From an operation perspective, being able to manage your database like you manage your application is something that is very interesting. We are not there yet. I mean, for us, the database is still outside of the OpenShift cluster but it's something we are working as well with Red Hat because we are using CouchBase, the data store layer and we are working with you guys to try to integrate that into OpenShift because it does make sense. We want to be able to manage everything the same way. It makes operation much easier for us. So, Rob, you're in the technology vendor space same as a lot of the folks here. One of the challenges with platforms is platforms in some parts of the industry tend to be a winner-take-all type of thing where you start with something and you continue to add features and you continue to add features and the ecosystem sort of goes, what about me? Like, what are you doing? What are you seeing in the Kubernetes space in working with some of these open communities that as a technology partner slash vendor you feel like you can continue to do what you do really well and not feel like this platform is going to sort of consume you, if you will. Yeah, that's a great question. I think applicable for us because we see a lot of customers. We go into them, they're not green fields, they're brown fields, it's messy and nasty and the first thing they come up with was what about all my batch workloads? I've got this 2000 node cluster running batch work and I've got some Spark apps over here and Hadoop workloads and I've got my container stuff and it's not a real clean cloud.com type scenario so you've got to integrate all these things and I think that's where a lot of the enterprise vendors like Red Hat and ourselves that have that experience on sort of pulling this brown field together and making everything work together and all day long we deal with that enterprise complexity whether it's trying to integrate mesos frameworks or batch workloads or all these non-containerized things that get really complicated. So, Shay, talk about it. You guys get probably as close to any customer as any. When you're an SI, when you're quote unquote partnered in the channel, you live with these customers every single day. Like, what does their journey look like? I mean, realistically what's it looking like for people because it's great to say, hey, new shiny thing comes out, you deploy it, person's a rockstar, everybody gets promoted, billion dollar growth but what's it really looking like for customers? So what are you seeing, they're bottlenecks, what are you seeing things that are making them successful even in small parts and what can you pass along to people? So the customer landscape that we've come to work with, typically the journey starts with one particular group and it's actually the group that we as infrastructure and ops guys wanted to go talk to which are the developers, right? Self-service was something that was very difficult for them to win in the virtual machine world and I should caveat this all by saying I am from Canada so our adoption of large scale clustering systems and self-service in containers has actually been a little more laggard than the United States. So we get introduced to a group and we say, new shiny thing, come check it out, do a POC and the developer experience is the thing that they latch on to, it is, wow man, I could really spin this up and get going very, very quickly, wonderful. That's not the end of the journey, that's just the bait and switch, right? Like that's, get them hooked on it, they wanna do the product and then you're gonna have to go and implement which usually means nine or 10 different groups of people that they haven't figured out how to work together with. So this is the traditional DevOps problem, this isn't the tooling problem of something like OpenShift, it's helping them out. That process, we've got sort of a couple of the customers in the room I've worked with typically takes them about a year where they start to say, new shiny thing, POC, develop a project around it and then wrangle ops and developers and network guys and everything else for the next few weeks, right? I'm working with one very large telco right now that is I met them in September, they'll be ready to start deploying in December and they've had an OpenStack platform available for that entire time, politics, right? It still tends to happen with most of our customers but at the end of the year, once that developer can go and do a demo and show live that they just pushed a new piece of code and it went into production state, that's the beginning of the next journey. And so the beginning of the next journey for us is to say, we typically give them a little bit of guidance and help along the way, it's check back every few months, let us know the problems you're having but it's also their own development journey, it is their own tooling pack, their own new CI CD tools that they haven't been able to play with before, their own, how do we monitor or manage this, right? That's the biggest question that we have today is like, how do I see what's going on in the environment, which is still difficult. And so then that next journey for the next year is always productive, like they can always get something out at the end of the day, they get their minimum viable product but then they start developing on the next thing and to me that's the fun part because I wanna learn where they're going and help them out along that journey because that's business value, that's not some of the technology problem, right? So there's a large part of this, this is own your change, you've got on it. Any tips that you've found to get, I don't know, a group to showcase to somebody, hey, this is worth your time even though the numbers might not be simple, like telling somebody they're gonna become more agile sounds great until you have to put it in a spreadsheet and have to somehow justify it. Any tips from any of you as to how you sold that internally? I would say the lucky thing right now is that if you're going from top down there's somebody's job is to make that happen right now. It's on their MBO list. So figure out whose job that is to say I need to get my teams working better and faster together that gets paid on that and then take each one of the groups and show them one demo and make sure that demo is what's in it for them. How is ops better? How is development better? How do I get the product out at the end of the day? Those are the sort of things that we tend to showcase. And definitely what is important is to start with a small project internally and not to try to address everything at once. I mean really start with a standalone application that is already more or less ready to be cloud, cloud ready and that's the way we did it. Really starting with a small application, a smaller application from our side and showcase how better it is compared to what we have currently so yeah. That makes sense, Rob. Are you gonna jump in on it? Okay, good. So this is a community-based event. This is a platform-based event. What do you guys wish we were doing better, right? So you guys have had a chance to sort of, this is great, this is positive. What do we need to do better? As a community, as a core set of technologies that you're hearing the market that's just like, why aren't you guys adding this? Because it's the most obvious problem we have. What do we gotta get better at? I think maybe we listen to ourselves and talk to ourselves too much as an ecosystem and not enough customer. It's great that Kristoff is here and it just seems like a lot of vendors still in the room and we still have that. That's partly just the state of the industry right now. But I'm thrilled with what I'm seeing as far as now that we're getting out to customers and we're running into OpenShift in the enterprise because we deal with Fortune 500 type of company. So we're seeing a lot more production than I thought was out there to be frank. So we're thrilled about that. I'd love to hear back from a lot more of that as to the successes, the use cases, the issues. And I think that just takes time. It's not a fault of anybody, it's just where we're at. What's coming? Yeah, I think based on the customers I've worked with the documentation and the number of examples and the ease of installation. All of those have really greatly improved in the last two years, no question. But they still can be definitely, those are target opportunities because a lot of people, they'll just say, well, how do I install OpenShift? And you're like, okay. So there's a side of it, before you can even sell them on the advantages of the technology, there's this hurl of how do you get them the right information and how do you make the installation process for them something that they can deal with? You see, so they don't just totally ride off the technology without. So I would always say that just continued improvement on the documentation, example sets and things like that would be hugely received by anybody using this. Yeah, let me ask the audience a question. For those of you that are evaluating OpenShift have played with it, how many of you know of this thing called OC cluster up? It's getting better? Okay, so for those of you that don't know, Kubernetes fundamentally wasn't the easiest thing in the world to get working. OpenShift wasn't the easiest thing in the world to get working. And in classic technology fashion, you can go and search the blogs. We now have about four or five different ways that you can have one command to bring things up. OC cluster up being the simplest one. It's gonna bring up an environment right in your laptop. You don't have to mess around with a VM or anything that's gonna bring the whole thing up. You can run demonstrations and stuff like that. But there's a whole bunch of others that take advantage of what's going on with Cube Admin, Mini Cube, Mini Shift. A bunch of things like that. Excuse me. So we're gonna be at CubeCon tomorrow. We're gonna see Kubernetes, we're gonna see Prometheus. Obviously Docker's a big piece of this ecosystem. What else do you wish from open projects that you know are out there was more tightly brought into this? Or what's on your list of things you'd like to see us bring in into the community? Sorry, so from my side, the hardest thing that we have right now or the biggest challenge that we have is around traditional monitoring frameworks have not yet adapted to container engines and a lot of them haven't sort of met up to actually go into Kubernetes and get all the data that you want. And metrics and logging are great inside of the stack but I think we need more than that. So I've tested out SysTix, got a great cloud product but they also have an open source tool. Something along those lines to say and to be more prescriptive of this is now the best way to monitor OpenShift. I think is one of the biggest and most common questions that I have and we can say, I can go to Nagios and I can build a bunch of plugins to go do that but I think while OpenShift strives to be open sometimes it strives to be too open. So maybe a project like SysTix or something like that seems to me would be interesting if that was included in the base package because you have an open source component. Yeah, I 100% agree. It's basically our main pain point right now. I mean, we have to get in production with OpenShift in the next, I don't know, six months and right now we are still struggling at knowing, okay, what type of monitoring platform we should be putting in place. If we were running on Google we could use a stack driver or this kind of stuff but here we are running on our in-premise OpenStack and we don't know, I mean SysTix could be an option. We, you know, it's something that would be extremely useful. So this isn't so much a request for, we want you to only ship us the SysTig new relic agent but we want you to give us some sort of guidance and if that guidance is around Hocular or Prometheus or whatever, just make your lives better. Yeah, yeah. Okay, all right, fair enough. What else, anything else on your list of areas that you wanna say, I mean it was the big data area that we talked about, is that an emerging space you're seeing from companies and customers or is that still sort of unique to certain industries? For me, not specifically big data but more really the how to manage the data store layer. So you know, the persistency, how do you manage the persistency? That's something that I don't think anybody has really sorted out right now and that's also a missing piece. As I said, we are looking to try to manage everything the same way on an OpenShift cluster and the persistency layer is one of the big question right now. So that would be the next. Yeah, I totally agree on the persistence. I think it's improved a lot in two years but things like being able to define what kind of storage you want just from a quality perspective, making it more transparent to the user so that they don't have to worry so much about the underlying back end storage and then also things like if you wanna have data replicated in your environment, that would be great just to have that baked in and transparent to the user. That'd be a killer feature to have added. To switch a little bit from the op side to the user side, one of the conversations I was actually having last night was surrounding maybe we, so the developer experience is great, a developer can go in and do something that's not really the same experience that we're matching up with the application owner or the team to say what do I have? How do I manage it and where's my community around supporting that? As I can deploy an OpenShift cluster, maybe it's interesting to say with that I also deploy these chat rooms and these integrations and these other sort of DevOps-y type tools that could be integrated to say, yeah, to say here's how I get help when I'm doing this because the first thing everybody's doing when they go in is learning. So broaden the experience, it's more holistic. Yeah, exactly. Gotcha. Gentlemen, thank you so much for this. I think Diane's gonna open the floor to questions. We're gonna cut it off. Can I get a couple of questions if there are? Yeah. Questions? Cool. All right, gentlemen, thank you for your time. Thank you. Thank you.