 Who here knows what Quay is? Raise your hands. Excellent. Who here is ready for lunch? Excellent. So I'm going to try to make this as long as possible. No, I'm kidding. I'm going to try to go quickly here. I'm not going to keep too much of your time. Five minutes max. So I'm going to just go through this. My name is Joey Shore. I'm the former co-founder and now tech lead of the Quay team. And I'm going to just go a little bit into the history of Quay and some of the cool stuff we're announcing today. So what is Quay? Quay was the first enterprise grade container registry ever. We actually launched Quay. Oh, this is an example from early 2014. We launched in October 2013 at a doctor meetup in New York City, actually the same doctor meetup where we would meet Alex Pulvy, who would eventually acquire us later on, which was a cool little factoid. And we were actually the first private container registry on the internet, even before Docker Hub. So when we launched Quay, you can see here there was a pretty big focus on secure hosting for private Docker repositories. And this was because at the time, there was no other solution for private ones. There was only the Docker public index. We quickly grew from there. And by grew, I mean customers, not us. It was just myself and my co-founder. So it was a two-man job for a very long time. And then in the summer of 2014, we were acquired by CoreOS. And we grew with them since then and grew out the Quay team. And then we were acquired by Red Hat in 2018 and then IBM in 2019. I've had the same job for four companies. It's actually pretty fun. So this was Quay back in early 2014. And this is Quay today. Not much has actually changed. It's actually the same code base that has grown up organically over time. And on a mark of that code base, what is Project Quay? Well, Project Quay is the open sourcing of our entire code base. 9,900 commits, of which I wrote about 6,000 of them if I had to guess. So it's been a very long time. We're looking at over six years of development effort now. And as per Red Hat's commitment to open source, Project Quay is the open source project behind the Quay container registry. I should mention the source code that we open sourced and that is Project Quay today is the code that runs on Quay.io and is the code that is Red Hat Quay. There is, as of this moment, no upstream versus downstream. It is the same repository. If you look at the PR list, you'll see that I actually have an open PR from Friday of an in-progress data model migration that is going on in our production right now. And that PR will be merged once that migration is complete. And you can just follow along as we do stuff there. And this is part of our commitment to trying to ensure that we keep as close to the source wherever possible. Both for our hosted product, which today serves hundreds of thousands of requests a minute and it serves literally billions of images all the way to our on-premise solution, Red Hat Quay, which is, of course, our managed solution that you can use with OpenShift. So the important parts, sourcing, distributing, the Project Quay source code is hosted at github at github.com slash quay slash quay or qq. Please take a look. We did do something a little interesting. We squashed the commit history because there was a lot of commits in there. There were just myself and my former co-founder like railing against random things. So rather than having to go through and hand-edit 9,000 commits, we cleared the history but we kept the empty commits so that the people who did contribute, even when it was a private product, were still there in terms of recognition. And then since then, of course, we will have full commit history. We also have Claire. So who here knows what Claire is? Excellent. For those who don't know, Claire is our open source security indexer and scanner. It scans for package-based vulnerabilities found in container images. Claire can be found at github.com slash quay slash clare. It works tightly with quay, but it is not required to use quay with it. So there have been a number of integrations of Claire by other providers. Amazon, if I recall, launched support for this. I believe there's no more than two weeks ago. And I believe there's been some work done by Google as well. So we have a lot of contributors there as well. And there is a new version of Claire, Claire v4. Yes, we have currently Claire v2 and we're working on Claire v4. There's history there. But Claire v4 is currently being developed as the more modern version of Claire. So we're always looking for interested parties there as well. Documentation for Project Quay can be found at docs.projectquay.io. You don't need the slash welcome but it's a good entry point. I should state for the record, these docs are evolving over time. They started as a duplication of our Red Hat Quay docs a couple of weeks ago and our docs team is working diligently to spruce them up and change them. So if you see references like rel is required to run Red Hat Quay, for Red Hat Quay that is true, for Project Quay it is not. So don't let that be an excuse not to run Project Quay on your own laptop. As long as you can run a container, you can run Project Quay because we serve it as a container image. There's a Docker file in the root you can just use to build. Some things I should talk about. I'm gonna put a big asterisk on this. Anything on here is speculative. Me is subject to change. I am not the product manager. So if my PM emails me after this talk and says, hey, why do you commit to this? I'm not committing to anything but these are the things that we're likely to do in the near term future. The first big one is code cleanup after recent data model migration. For those who don't know and I'm gonna go very briefly into this but there have been three different versions of the Docker protocol for pushing and pulling container images and Quay today supports all three simultaneously full backwards and forwards compatible. We're the only registry product on the market that has done so and will likely continue to do so. But in order to migrate between the older APIs and the newer APIs, we had to migrate our data model and we had to do that on Quay.io with a billion plus images without any downtime. This is like trying to replace the engine and the wheels of your car without stopping on the Autobahn. And so I intend to give a really cool talk about this in the future. There's a lot of really cool history here and a lot of very subtle technical detail. But suffice to say there will be a PR coming for me in the next two months or so that's going to delete a very large section of the Quay code and I'm very excited about it. So that's number one. The second one is additional modes for registry mirroring. We should actually say repository mirroring but I mistyped it. Quay launched a feature in Quay 3.1 which went out a month or two ago which was repository mirroring where you can now set up a Quay repo to mirror the contents or a subset of contents from another repository. I also need to learn to speak English whether within the same registry or another registry. So you could for example have your on-prem Quay mirror images from Quay.io or Docker Hub or another on-prem Quay. You could have a production namespace mirror, a subset of tags from a staging namespace, things like that. We're going to be adding additional modes and additional rules for registry mirroring, repository mirroring. Right now it's by tag regex but we're going to be adding label matching and a few other really cool things to allow you to have very fine and powerful green control. New build manager support. Quay has built-in support for building Docker files. You can give us a Docker file via directly uploading it or hooking it up to your GitHub or GitLab or BitBucket or just even your custom Git and call us up with your web hook. The build manager for that on-premises uses just build workers connected to Quay but as the world adopts Kubernetes, we have a Kubernetes-based build system that we actually use for Quay.io and the plan is to formally support that for OpenShift customers. So you can just hook an OpenShift cluster to your Quay and it'll use OpenShift to do the builds. And then finally, and this is the most speculative but also the most fun part in my opinion, the Quay API was originally designed when the world was a little less complicated and was a little less based on manifests and more based on images. We've added manifest support to our existing version one API that Quay continues to support today. However, the plan is to build a new API, one that is faster and doesn't have some of the cruftiness of the older versions and to hopefully revamp our UI to make use of that new API for better performance and efficiency and additional feature gains. Quay is one of the few registry products on the market that is fully automatable. Every single operation and action that is possible in the UI is driven by a fully documented OAuth-based REST API. And so this will be the next generation of that and we're always asking for feedback on it as well. So that's some of the upcoming work we have. There's quite a bit more. I would suggest taking a look at the Red Hat Quay roadmap as well. Next month is likely when Red Hat Quay 3.2 will be launched and there will be the very first integration between Quay and OpenShift in our container security operator work, which will allow you to automatically be installing the container security operator in OLM, see the security status of your running pods in your OpenShift console if they were pulled from any version of Quay that is connected to a Clare with no additional setup. There will be no off-token management, no connectors, it'll just work. So I would recommend taking a look at our roadmap there as well and that'll be a big, nice feature coming up. Don't forget to start our repos on GitHub. We have 1.1 thousand since we open sourced late last week or middle last week. So for less than five days or five days, we're getting good traction but we can always use more. We need to get that number up and with that, I would say bon appetit and enjoy your lunch. Come up. So that took a whole lot of work to get to there and I just wanted to quickly thank Joey and the whole entire Quay team and I know Joey doesn't drink so we have a glass of water for you but if you could join us, Joe and we could have a toast because this took a lot of work and come here and... I haven't come up here. Evan was previously part of the Quay team. All right, get Evan up here. We're gonna move over here. We're gonna pose, hey. Okay, this is the first time I've seen Evan in the flesh. I've seen him on Slack. Hi, Evan, here. I didn't mean to put you on the spot, Evan, but... There you go. Thank you. That took a lot of work and we totally appreciate all the effort. Thank you very much. Cheers, guys. APPLAUSE Perfect.