 Hi, welcome to SuperUserTV. Can you go ahead and introduce us and what you do at Paddy Power Betfair? Hi, I'm Stephen Armstrong, Principal Automation Engineer, and at Paddy Power Betfair I'm the technical lead on that I2 project, which is our next generation OpenStack infrastructure. Awesome, well congratulations this morning at the OpenStack Summit keynote. You were, you tied for the SuperUser Awards with UK Cloud, but Paddy Power Betfair was a winner. So can you tell me what you do with OpenStack at Paddy Power Betfair? So at Paddy Power Betfair we run two OpenStacks per DC. We run an active-active infrastructure. What we're doing at the moment is we're in the middle of a migration project. We're migrating all of our customer-facing applications onto OpenStack. So we're migrating around 25% of the applications at the moment. And we're going to ramp that to 100% and hopefully migrate it in the next year. And I know one of the stats that was really impressive in your application was talking about how many deployments you do a week or how many do you do a day. So can you kind of talk through that process? So what we've really tried to do with Teams is we bring them in for a sprint where they will onboard their micro-service applications. We try and create a self-service model, and so that we teach them how to use the OpenStack infrastructure. We've created a self-service model with Ansible where they'll fill in self-service inventory files, check it into source control, and then they can basically onboard applications themselves. So the first time they onboard, it takes a bit longer because we're teaching them how to manage the infrastructure themselves. After a while, they just manage to deploy the applications themselves. So that's really the key here is basically training them to be a full self-service model and basically not to be a blocker for the developers to get their new products to market. And so with all these different processes and the scale that you all are working at, one of the pieces of feedback we get is that you have to have a large team to run OpenStack, and how do you feel with your team and what is your overall structure? So what we've really tried to do is control everything programmatically. So we've made all the processes repeatable. So we were going to run at the scale that we were doing this all manually. It would just be impossible. So what we've got is we've got a main core team of around eight engineers that look after the main OpenStack platform and our software-defined networking with Nuage. They look after the upgrades and basically make sure that the platform is performing. We then have onboarding teams around each location. So we've got one in the UK, one in Dublin, one in Romania and Cluj, and one in Porto that helps all the development teams to onboard onto the platform and migrate their applications. So they're small teams and because everything that we do is automated, it's really just an educational piece to try and get them to use the platform and get used to doing it. If we have any features, then the core team and the onboarding teams will implement that new feature and wrap it in the onboarding process. Awesome, and so we're here at the summit. What are y'all talking about? I know you have, I think, three different sessions that y'all are sharing different experiences that y'all have. Yeah, so some of the new things that we've done is we're onboarding our Inframix databases and MySQL databases. So we've got a storage chat on that, on how we incorporated those features into our deployment pipelines. We also have one for shared file service. So we're using the Manila project for that. So there's a session on that and we've got a few live demos for that as well, which should be good, hopefully they go well. And the other main one is really a lessons learned of running 1,000 deployments a day on OpenStack. So we'll take users through all the problems that we basically had and how we overcame them because you have lessons learned when you create infrastructure at this scale and we just want to kind of share some of those experiences back with other users so that they can learn from them and not repeat the mistakes that we made. Awesome, well thank you again for chatting with us at the Boston Summit and we hope to see you again soon. Okay, thanks very much.