 I'm going to bring out from IBM Brad Topol. Hey, good to see you, Brad. So I mentioned that what you've been working on is taking an application and showing how to deploy it in a standard way across multiple open stack environments, right? And you've had someone helping you with this that's going to come demo this today? Yeah, I brought a few friends today. OK, all right. Well, why don't we go ahead and bring out the demoers for this? And we can get started with this demo. It looks like more than a few friends, Brad. Yeah, we had a few extra folks who decided to crash the party. So you guys can go ahead and get on your laptops, but don't start yet. This is something that we've been calling the interock challenge. And I know that one of the people that really kicked this off was Don Rippert from IBM. And he's going to be speaking in a minute. But since he's here, Don, maybe you can just come out for a minute while they're getting set up. So Don, I know that you made this challenge because you were feeling like people were missing the boat on the real story around open stack interock. And now that you see this, what do you think? Is this what you were expecting? I think I should index Brad's bonus to participation. I'll call the CFO and give him the bad news. I'm going to say that. Absolutely. The people doubt that open stack creates interoperability. There are doubters out there. And so in April, I said, let's prove the doubters wrong. Extended an offer to our competitors, our cooperators, our community members who are vendors, and said, let's see what we can do with interoperability within the next six months. Let's bring it up on stage in Barcelona. And let's prove whether this interoperates or not rather than just talk about it. Yeah, OK, cool. Well, I think this is going to be really interesting. I've never done a 16-person simultaneous live demo before. But we'll see how this goes. So thank you, and we'll see you in a minute. All right. Everybody ready? All right, start your demos. So while they are kicking this off, I want you to explain for a minute what we're talking about. Because this isn't just about taking some code and deploying it on a server. This is really about an entire architecture, right? Yeah, it's an entire enterprise workload. We had a couple goals with this challenge. We wanted to demonstrate one that we were, open stack was able to deploy an enterprise workload, something that had a load balancer, had a database, attached volume, had security groups to control the network behind the scenes of the different nodes. And it had WordPress nodes and a database node. So putting it all together as an enterprise workload was key. And what we wanted to show is that all these different folks could be able to automatically deploy this workload. So we looked at using Ansible, and we looked at the shade module of Ansible. And these vendor-neutral tools gave us the ability to deploy this automated workload across all these different clouds, the private, the dedicated, the public, as well. So between these different deployments, what is the variation in terms of what they're deploying, what they're running? Yeah, they're all running the same Ansible scripts. Each one uses a little bit of a config file, because you have to say, hey, what's the endpoint for the different cloud? What's the user ID and password? And there's a few other minor values that are slightly different between public private clouds. So if you go look, and what's nice about this work is all of this is available. This is in the OpenStack Ops contrib repository. All these folks work together in the open to do the best practices for how you would do an automated deployment of enterprise workload. So everybody can go out and see the work that they did, and see the very small piece that they have for config. But all the rest is the standard scripts that they all run. OK, let's talk to them. I've got my game show mic here. So we're in Europe. Why don't we start with Sousa? What have you got happening here? So the deployment is actually moving really fast. So we've went already through the prep stage. And actually, the virtual machines have been created already. OK, is there anybody who's creating virtual machines right now? Oh, OK. Got some down here. So why don't we go to OpenTelecom cloud? Yes. So currently, the balance of notice being provisioned. You can see that here in the script. And you can also see that in our console, where the database note is already there, and the balance of notice is in creating. OK, cool. So Cisco, what have you got going on over here? So we got all the instances up and running. And this instance is also a little different that we have a provider now we're setting. So all the virtual machines are getting IP addresses which are internally routed within the data center. OK, I see that. So we all got that going. And we are installing the MySQL libraries into the VMs themselves that have been provisioned right now. OK, cool. And canonical, what's happening on your side? So we're pretty much at the same stage as Cisco. We're installing MySQL and associated libraries. And we can see that the instances are up and running. They've been got their floating IP addresses. We've got our block device allocated. And we're just running through at the moment. OK, so sounds like you guys are in a little bit of a race here. We'll see who finishes first. So provider networks, anyone running tenant? We'll go to Tong here at IBM. How's it going? Good. Can we see? Yeah, OK. So you've got floating IPs here. Yep. So right now, my script already finished installing MySQL and libraries, and doing some other stuff as well. So you can see that they're from Horizon. Right now, I just showed, before I start the script, if I refresh it, I should see all the VMs provisioned already. You can see the database node, web server node, and load balancer. And each of those nodes should have the private IP and the public IP as well. OK, cool. So you mentioned security groups earlier. Is there anybody that has security groups that are visible that they can show us how that's been? Craig, you from Arantis? Hi there. Yeah, we can show the security groups. So there is a security group automatically set up by the Ansible script. So it should look the same for everybody. And this was automatically created through the OpenStack APIs. And it's exactly the same as in every other cloud. The thing I like about all of this is it really just isn't that exciting, because it works the same for all of us. It works the same for all of you. That's actually an interesting point. So we have two of the participants here. We've got Agla from Rackspace and Mark from VMware. And where are you at in your state at Rackspace here? Actually, I just completed. It says that I already have a VM. And the WordPress site is running. Just refresh to make sure that it's not a cache version. And I guess it takes a while to get to the data center itself. But yeah, it says that I completed in a little under five minutes. OK, cool. So Mark, one of the things that is really important about this is that it's using the standard set of APIs, even though there are differences, perhaps, between network topologies. And like Don and Brad were saying, those are the things that sometimes people go, OK, well, this is what prevents OpenStack clouds from working together. But in your work on the Interop Working Group, you have really increased the coverage of those API tests. And this is sort of a functional demonstration of how that works in practice. Can you talk a little bit about how that is in place here? Yeah. So as part of the Interop Challenge, we actually asked all the participants to run the RESTAC suite. So they've all submitted tests up to RESTAC. The one we're looking at here is for VMware Integrated OpenStack. And you can actually see the results here that this actually complies with the most recent guidelines. And you can get a list of all the other things that we do. The Interop Challenge workload itself, almost all the capabilities there are part of the DevCore guidelines. OK, so basically, there's a really good chance that you could take this one set of orchestration scripts and run it on anything that's passing the OpenStack powered. All right, so we've got a lot of diversity up here. I actually asked people where they were from yesterday. And I think we have 11 countries represented. In addition to all of these different organizations and Red Hat, you are? Well, actually, I'm from Madrid, Spain. Yes, I'm from CS2, or attendees. I put horizon here in Spanish. So you can see here the network topology. So we have all the servers bound here. And you can see that the website is already up there. OK, great. So I guess that that's basically it. So this uses shade to respond to all the servers. You can see all the funny things here on FloodNet. But I guess that's a real good attempt, yes, to see how this is interpretable. Yeah, I mean, this is the network topology that's generated automatically. And one final mention of continuing the diversity at the technology level, Lenaro, what processor architecture are you running on? We're running on ARM64. OK, ARM64. Yeah, so if you see the balancer here, I can actually connect to it and show you. OK, so yeah, ARM64 over there. So this is really cool. So I think that you were able to accomplish your mission in taking a not just, again, code, but a network architecture, a deployment methodology, security, load balancing, all of this, and show that it can be run in an incredible variety of open stack environments. Yeah, there was a holistic approach like operators would need to be able to show interoperability and have that good experience. But I mean, we've got SUSE and Red Hat and Canonical. And I mean, these people don't like each other, right? All right, these were the best folks to work with. Everybody worked together on best practices. Even at the last minute, my Cisco friend didn't have his dongle to connect. I gave him mine. We're all working together here because we know the real winner here is not who's fastest. The real winner is the deployer, the customer, because everybody's focused on interoperability. All right, so how many of you were able to actually get this workload running in your different cloud environments? All right, cool, 16 out of 16. So thank you. So we actually made a little plaque to commemorate this that says OpenStack Interoperability Challenge 2016. And that one's for you. And I'm not going to try to hand out all of yours right now, but we have them backstage. So thank you guys for participating. And this is a really cool demonstration of how interoperability actually works, how this type of multicloud can function and do it across an incredible array of systems. So thank you. Thank you. Thank you.