 All right, everyone welcome to a spinnaker summit at CD con 2023 We're co-located here with CD con and githops con and there's also the open-source summit going on I'm Cameron Moda boss Elani. I'm an engineering manager at armory and a spinnaker to see chair This is the state of spinnaker So first I wanted to go over the project history. I gave a presentation of this to the CDF TOC a few weeks ago and thought it would be good to just go over that with the community as well So spinnaker was first created by Netflix in 2015. You can see it commits back back all the way back then The project was open-sourced After Google joined the project Over time up armory and op-semex joined the community and started to participate and really make Spinnaker the open-source community is it is today? Oh We were participating yearly in the Google summer of code and we were hosting week-long community hackathons And then COVID happened. We had some major contributors and community leaders that that left the project Contributions went down the steering committee and the technical oversight committee were not as engaged as they were in the past The release process had some issues. It was broken, but we had Salesforce joined as a major contributor We took a look around and we said hey, there's still a lot of people using spinnaker today large companies large enterprises And there's a dedicated community to spinnaker. So we said hey, let's get together and revive the community So we started off by meeting in person in Austin 2022 at CD con about one year ago we recruited a variety of new technical oversight committee members and We went down the list and we we looked up major contributors Vendors and then some of the largest users of spinnaker We wanted to have a good variety of folks to show that there's large companies Vendors etc dedicated to the project. So we went with nine members total This wasn't done in isolation though We worked with the previous technical oversight committee steering committee and the CDF on updating the governance structure We combined this technical oversight committee and the steering committee into one group and the intent there was to make decision-making more agile before I talk about the motivations and how we we got together to to get alignment on the community direction Just wanted to talk about the talk. I saw yesterday Robert Reeves gave a really good talk about how to show business value to your organization. How do you affect change? As engineers, we see the value of continuous delivery, but it's hard to bring other folks along along the way so Just wanted to shout that out because I saw a lot of Comparisons and parallels with the work that we did of reviving the community So we talked with all these members that that were interested in joining and participating and we gathered motivations for joining the technical oversight committee We then used those motivations for setting expectations for TOC members And for lining on the project roadmap Some of the top motivators included increasing contributions That's the lifeblood of any open source project is the contributions But other motivators included using more open standards reducing breaking changes having a really good stable product And improving on the release process Right on. So let's recap what we did in 2022 Well, we fixed the release process Big thanks to David and Carl Scoos for for working on the release process and bringing it to fruition There's still more improvements to come and we'll we'll go over some of those later We're also setting up the release manager rotation and now it's not just One company doing releases anymore. We've got at least three of us and more getting onboarded as time goes on We also restarted the rfc process rfcs are a request for comments and that's how you affect change in the community You put out your idea of how you want to affect or impact the project And then get comments on it and then we collaborate on the the right way of implementing a new feature So three rfcs have been approved since we restarted this process Those include the mono repo. So we're going to be going towards Consolidating our repos into a mono repo CD events as well as backporting changes behind feature flags These rfcs have already had a really major impact on the community development All right, so let's go over what we've done in 2023 so far So we've been improving contributions to spinnaker. We're reducing the barrier of getting entry for getting started Carl scuse has been working on and has donated the spinnaker customized repo and that's a great way to install We're also working on modernizing spinnaker. This includes upgrading java spring and gradle This is done for a few different reasons modern language Features are great to use It also helps improve security keeping up to date with the latest patches Make sure that you you have few cds within your your code base And then we're working on the mono repo This is going to help simplify both development and release spinnaker is comprised of 10 or 11 microservices and we have a common library as well The dependencies for this is not trivial If you make a change to the common library, it then propagates to the other services If you make a change to orca, it propagates to cloud driver If you make a change to fiat, it updates a few other services As you can see this causes a lot of Of changes and it's really hard to keep things aligned By moving to a mono repo, it's really going to help simplify and make sure that our changes are correct when we make them across the project We're also working on cd events integration That's that was one of the rfcs that was Approved and merged and we're working on triggering pipelines based on cd events coming in Next steps for that though is to become a producer of cd events once we produce cd events We can start to join this ecosystem of of projects that integrate with cd events And then performance improvements we've had some pretty large performance improvements In the in the past and we're going to continue seeing those as time goes on Right and now let's talk about some industry trends that we're seeing today So first I just wanted to go over this idea of crawl walk run The way that a lot of people get started with continuous delivery and continuous deployment on their digital transformation journey, if you will A lot of folks are are doing manual deployments and they they see all these shiny tools out there and hear terms like declarative delivery imperative delivery And and so on what do all these things mean? It's it's confusing when you're first starting out If you go from manual steps directly to uh declarative that's going to be a pretty hard transition so I definitely see this trend of crawl walk run really Playing out in companies today So there's a case study I've linked here and there was a Company that automated the manual steps They had a 250 percent 256 percent return on investment in that What did that do it gave them time savings and it also improved the correctness by taking manual steps out you're you're reducing the Possibility of issues This also helps your engineers focus focus on different types of work There is a real opportunity cost to having your developers spend time manually deploying and having to roll back and understand your your infrastructure Any issues that might come up that's now time taken Back that's that's time taken away from your your developers and they also have to fix it So that really increases the time taken to deploy a new feature so Let's take a look at What you can do with the time you have saved once you automate things So this is another trend that that I've been seeing and it's using the right tool for the job When you have the time to invest in in Your organization and your your capabilities you can now use the right tool for the job So prime video posted an article recently where they moved their from their lambdas solution to ec2 and ecs They saved 90 of their infrastructure costs doing this This has not just cost savings, but also a great environmental impact We're no longer. They are no longer shoving bits just for the sake of moving bits around And this is not at all to say that their choice of lambda was a bad choice. That's really what helped get them started It's easy to get started and they were able to run fast with it And just to show it's not just me thinking this here's a quote from help kelsey high tower He's saying, you know, this isn't to dig against lambda at all there's a quote within the the article that prime video posted about How it was a good solution for them But they had to migrate deployment targets. They weren't they were deploying to lambda at first and then they're deploying to ec2 and ecs Uh, how do you do that with one tool? Well spinnaker can do that Another trend that we're seeing right now is is open standards So there's a few different success stories that I'd like to highlight. There's an open telemetry project Which is gaining a lot of traction There's also the service mesh interface spec, which has drastically improved the the way service meshes integrate today We also see a lot of users Leaving twitter and going to mastodon people want to take control of their own Their own data and have the ability to to be a bit portable And I'd like to go back to the idea of cd events. This is a spec that's based on cloud events And that comes from the cncf CD events are a subset that builds on top of that And I would encourage everyone to check out the cdf events special interest group Finally, we have a we see murders and acquisitions on the rise So cart to release to report the q2 q1 2023 investor landscape How is that relevant to spinnaker or cd in general? Well, a lot of us work at large large companies and large enterprises Um When you acquire a company, you're not just acquiring the people you're acquiring their tech as well And while it might not happen overnight, you do eventually integrate that tech and into your uh into your central organization There's a lot of Benefits for doing this. We talked about cost savings earlier But this also really helps with compliance if you have a really fragmented Way of managing your infrastructure that gets really hard To just understand and wrap your head around As a cto or cio you want to know what your infrastructure? You're running and paying for Well, spinnaker has the ability to manage all of that Again, this is challenging as companies are acquired and then your teams use new technologies Right now kubernetes is really hot But there might be some other tool that is the right tool for the job And if you need to deploy there, it's going to be hard if you're not using a tool like spinnaker So, yeah, spinnaker allows for many deployment targets and allows for managing and identifying the infrastructure you're using