 All right. Oh, we're starting. All right. Great. Welcome to the flux project lightning talk. Hey Yeah, hi friends All right, um and flux provides a secure and scalable get ops Yeah, I mean if you don't already know Tomo, this is Tomo. Oh hi. I'm Tomo. Yeah, so but yeah We're the here at the flux Flux lightning talk and we also have flagger, right? Yes, we put our names at the end. So I'm Tomo Nakahara I guess my hat is community manager xxx. Yeah, I'm Lee. I work in R&D it on Tanzu Excellent. Yeah, so what is flux and in parentheses flagger, which is a sub project So flux is really the project that created the term get ops, which the short term is Operations by pull requests. So basically flux really leverages the power of kubernetes It listens to your repo where you have a yaml file that defines the single source of truth of your cluster And then when it notices a change it tells kubernetes Oh, please make sure that the cluster looks like what's in the yaml file So if any bad agents on the outside try to make changes, it'll completely get overwritten and Flagger is also a sub project of flux that provides progressive delivery like canary deployments in between that process Which will explain a little bit more so it is a graduated project as you know, we also have general availability and Security scale and reliability are central to flux because many many enterprises such as financial institutions government many more rely on flux and many of the vendors such as Microsoft AWS GitLab all rely on flux under the hood to provide get ops to their customers So what is under the hood? Yeah, a bunch of a bunch of components and pieces that you can really depend on I love that you mentioned that we've made the push in the flux project to hit general availability because when you Consider being that level of production ready with the with the number of promises that we're making across all of our APIs and all of our features To say that flux is generally available is a huge accomplishment So thank you to anybody in this room has helped us either as a user providing feedback Or as somebody who has gotten our APIs to that finish line as you know contributor to the project Thank you so much flux is built with security scalability and reliability in mind we implement flux as a series of microservices several controllers own all of the individual pieces that you need to have a Sophisticated approach for your continuous delivery get ops, right? And so we have a helm controller if you are a helm user helm users anyone. Yeah Flux loves helm We love helm so much that we partner with the helm project to improve the helm SDK so that you can build a Controller around helm that does helm in a declarative way, right? So declarative rollbacks declarative hooks You know declarative drift detection all the things that you don't normally get when you are using helm on the command line tool use flux You can do the get ops native way for helm We also support customized so when you want to apply any of your manifests from wherever you are committing those two We're building them from from source code You can like modify them patch them use different environment variables and your different environments for your you know five ten thirty Five hundred clusters, you know, whatever your you know kind of fresh experience of Kubernetes is in your environment Flux is ready to scale to that Flux has an amazing bootstrap ux And so like the bootstrap problem is familiar to anybody in infrastructure who's trying to get systems stood up With their dependencies all kind of operating and then glued together to actually operate with each other And if you want to have a great bootstrap experience in get ops We integrate directly with git lab github and many other git providers The community has also been so active in building things like extensions for code editors like vs code user interfaces we've got actually multiple user interfaces out in the community for Using flux as a visual representation of your cluster so you can understand what's going on your developer platform and so much more get involved in the community Flux is multi cluster It is multi-tenant the way that we built the flux API is so that you could shape it to your particular Continuous delivery approach. It's going to meet your specific needs and we've got security in mind with that tenancy Tenancy examples and all that Flux can use multiple sources. We have support for you to just commit your stuff to git repositories You can also build your manifest package them into OCI repositories and use that as a source You can commit your yaml to s3 buckets and all kinds of stuff and then verify all of that with your GPG keys You could do cosine signatures for those very big environments and we have lots of flux users in incredibly large environments We even have implemented Sharding on label selectors for all of these resources so that you can split up the work Against multiple instances or multiple replicas of all the individual flux controllers. So security reliability and scale All in an extensible way that interoperates with many beautiful cloud native projects We also have flagger flagger is a sub project that lets you do continuous delivery And so if you're looking for things like canaries blue-green deployments a b testing traffic mirroring all with automatic rollbacks Extensible hooks to you know make make it so that somebody has to click a button in JIRA before the canary rolls out You can do that with flagger. It's awesome Even though JIRA is not sometimes right We've got ingress support gateways Service meshes the way that flux is built. You don't even have to change any of your configurations You can have your deployment sitting there add flux make a canary object and then suddenly it's rolling Like you don't have to make any special resources learning any new APIs It can integrate with Istio or whatever other thing you've installed in your cluster We've done a lot of work to support that The way that flux does this is that you have to it has to instrument into the metrics of what's going on with the Traffic flow of your application and it has to understand if your app is Healthy or not and so we've got support for Prometheus data dog many others with a bunch of default behaviors and custom metric support All right, so what's new here at cube con? There were no CVEs in our recent security audit as you heard We work really well with helm and even better with our latest release if you care about benchmarks check out our blog post on that and here at cube con We've got a flux maintainer talk where Stefan Pradhan will be talking about the road map And we have a wonderful birds of a feather tomorrow where we'll be showing all the ecosystem members contributing to the future of flux so Come by our flux booth in the project lounge We're also part of George's tour and flux CD.io slash cube con is our site Where it has all of our sessions and activities or scan the QR code, so thank you we will have custom stickers So come to the pavilion