GitHub Open Source Friday, 8 Oct 2022

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Preview features before merging with Uffizzi - Open Source Friday

Description

What if devs didn't have to check out a branch and run it on your local environment to test changes? Well, it's possible with Uffizzi. Watch the full video with Grayson Adkins to learn how continuous previews are changing the status quo.

0:00 - Introductions
1:27 - What is Uffizzi?
4:30 - Demo
17:38 - Answering audience questions
33:53 - Fun Chit Chat

Join @blackgirlbytes live for Open Source Fridays at https://www.meetup.com/github/.

Thanks!

Connect with us.
Facebook: http://fb.com/github
Twitter: http://twitter.com/github
LinkedIn: http://linkedin.com/company/github

About GitHub
GitHub is the best place to share code with friends, co-workers, classmates, and complete strangers. Millions of people use GitHub to build amazing things together. For more info, go to http://github.com
Chapters

A

Hello: everyone thanks for joining us you're, watching open source Friday, where I talk with open source, maintainers and core contributors about their open source projects. Today we're joined by Abby who's, my co-host and then we're also joined by Grayson. um Do you all want to introduce yourselves.

B

Sure hi I'm, Abby I guess I'll go first. Sorry.

A

B

Okay, I lead github's, open source maintainer program, so thrilled to be here on my first open source, Friday stream, and to hear from Grayson one of our open source. We dance.

C

Yeah, first uh stream, as well so I'm, Grayson, Atkins and I, am the co-founder and head of our product team at ufizi.

A

Nice, nice and I also see that there's, like seven people already tuned in so I, would love to know like where y'all are calling in from so feel free to like jump in the chat and just like tell us where you're coming from sometimes I see people from like Italy and Denmark, and all that so would love to know.

B

All right, finally, confusing that would yeah.

A

C

Look at our Italian, our Italian Brethren there right there.

A

I love it all right. Well, we can jump straight into what is you fizzy? If you don't mind letting us know well,.

C

We are not an Italian company or.

A

C

In the art art museum space, despite the name.

A

C

Is an open source API for creating and managing um ephemeral and event-driven test environments, and so what I mean by those two words ephemeral and event driven? Is um these environments are meant to have a predefined life cycle uh they're? They live they're they're born they live, they die and they have a very set purpose and are not intended to be persistent. Long running like your typical production or staging environments um and the second term there is event driven. That means um most commonly that's a pull request event.

C

So what can happen is, if you add an example of uh up easy, GitHub action. You can add that action to your pipeline, which we'll show here in a minute um and that's watching for events, live pull requests. So you can open a pull request on the feature. Branch that'll kick off a series of build and push steps building.

C

You know container images if easy can pull those images, create an environment for them, deploy them onto that environment and do some fancy networking magic and create a URL that gets posted back to your GitHub comment and then, whenever you can then take that endpoint share it. With your teammates run your automated or manual tests against when everybody buys off on it likes how it works. You can close that pull request.

C

That's another event that youth easy is watching for, and it'll tear down that environment, and so it's really about creating on-demand environments or in environments as a service capability. That's going to give you access to environments whenever you need them as long as you need them, and so they're really great for pull.

C

Request environments, like I just mentioned, also replacing your staging environment with these on-demand environments, you're using for demos, debugging environments, all sorts of things like that, so um it really gives folks a way to have more flexibility in terms of their their testing framework.

A

Gotcha and as I'm trying to imagine it more, would this be like a good option if you, if someone's like hey check out my pull request and like instead of you, checking out their branch and trying to figure out everything, you can just click the link and see what the results are. Yeah.

C

Absolutely so it's really a good tool for team collaboration, and what you said Roselle is is a huge part of that. Is you know it's not just the developer who's building the tool you've got. You know other developers who need to review the code, but also product managers, designers, business, folks, anybody who's, a stakeholder in your development process, giving them a way to asynchronously go check a link. You don't have to be online. You know to share your screen or anything like that.

C

You can, you can just send them a link and they can check it out when they need to and and all of it's integrated with GitHub nicely into your comments. Your pull requests they're there for, for whatever your team has time to review it.

A

Nice I'm excited: let's see it in action if you're you're, ready.

C

A

C

Sure so let me share here.

C

We go all right, so this is um an example application uh you can imagine. This is uh a team project. You might have multiple developers working on this and for all this examples, we're going to be talking about sort of a get branching strategy, so you've got some sort of main or or uh you know, develop branch that all your teammates are are merging to and then you're creating feature branches off this, this main branch- and in this case this particular application, is a simple voting application, cats and dogs.

C

You can, you can see uh right now the dogs are winning with with 16 boats. Here I can go over here. You can vote for cats and dogs, simple application, but in terms of its site complexity, it's actually um it's actually composed of multiple container services. So we.

A

C

A voting application, a front-end written in Python. We have a results: application, a node.js, there's a redis container cash, that's receiving votes, there's this net application, which is pulling those votes from the cash sending them to a database and so um well. The front end seems simple.

C

The the point of this uh is to show that you know this is a multi, multi-service or microservice based application, and so there's different components on here, both front-end components and back-end components, and so, let's say as a developer, I've got a new task and I love, dogs and so I want to I want to cheat the system a little bit and I want to vote for uh dogs twice. So that means every time someone comes in here and votes for dogs.

C

We want to give them two votes instead of one so that that's a that's a backing, change that we're gonna make to um in this case, the.net application- that's gonna! That's gonna, be doing that. So we'll demonstrate how to do that and how to kind of create a parallel test environment where we can check that feature before we ever merge it.

C

um So here I am in vs code and I've got this application open. As you can see here, I am on our main branch and so I'm going to go ahead and create a new um I'm going to create a new Branch for uh this double dogs ticket that I'm working on so I've switched to this double dogs. Branch and I have got a little code here that I'm going to add to it and I'm gonna say so.

C

Here, we've got uh if, if the dog, the second option, option b gets selected, we're going to have this uh two votes be added to it. Instead.

B

C

One, okay and so we're gonna, we're gonna rig this. This vote a little bit and uh so I'm Gonna Save that and then I'm going to see.

C

You see that I've made that change, I'm gonna, add it and then I'm gonna commit that and let's see so we're going to say double the dogs vote and there we go and I'm going to push this up to GitHub, where my repo is.

C

So there we go I've got it pushed up here and I'm gonna go check out my my um my my changes and see it's where they are so now you can see here. I've got this branch that I've just pushed up here.

C

I'm gonna go check that out and, as you can see here, it's one minute ahead of Maine I'm gonna go ahead and open a pull request for it and say: hey I think this looks good to go so I'm gonna create a new pull request, I'm going to select that double dog branch that I just pushed up and there we can see our change and so I'm going to go ahead and create that pull request and say, go and create.

C

And now what we're going to see here in just a second is a series of GitHub action steps and you can see them popping up and I'll talk about how we configure these in just a minute. But as I mentioned earlier, this application, it's got multiple Services. All of these. We have a GitHub actions pipeline, that's configured to watch for events like a pull requests, and so I've got a series of build steps that are building the the vote component.

C

As you can see here, the worker component, the result component, the load balancer and all these others, and so these are building, in this case Docker container images, okay and they're being pushed to a GitHub container registry. Now I can push these anywhere.

C

Actually I can push these to ecor or Docker Hub or wherever I want to keep my images so they're being pushed, and then here at the end of these steps, you're going to see a series of easy preview steps, and so what that means is upz is going to be watching for once these are built. It's going to pull these images and it's going to create a test environment for them and that's going to deploy those those application images onto that environment and generate a secure URL.

C

That's then going to get posted back to my GitHub pull request issue just below here and so that'll. Take a couple minutes to to finish doing: let's go ahead and check out how this is configured for those watch following along at home. So if you're not familiar with GitHub actions, they're typically stored in this dot, GitHub directory here.

B

C

Repo and you can create workflows- and in this case this is what's defining all those steps, all those jobs that are happening with it, that we just looked at so this this. uh These series of actions are going to be triggered on push to these branches. So if anybody is pushing these it'll build also with pull, requests are being opened, opened, reopen synchronized or closed so.

A

We're watching I have a question: yeah, I and maybe I should know this, because I work at GitHub, but I'm not familiar with the pull request, type of synchronize, and maybe somebody else isn't too yeah.

C

So so that has to do with- um and we can show this in a second, but um so once your application is built and deployed and you've got it at the secure URL. The other thing UFC is going to do. Is it's going to update that preview? Every time you push new changes, so even once it's deployed I, don't have to open up the new pull requests or change. The configuration I can just keep pushing to this Branch. That's on the open, pull request and that's what the synchronized is telling is telling GitHub.

C

Is that hey yeah we're synchronizing changes between uh what? What was there once and on those events, um go ahead and kick off this pipeline again.

A

Thank you beautiful explanation. Yeah.

C

So you can see here, you've got the we're building that vote application component and, in that step, we're building a worker component, we're building that results component, um the load balancer, which is receiving traffic HTTP traffic, and then we have a few steps here and essentially what this is doing is we are rendering a Docker compose file and I'll show you how what that has to do with this in a second.

C

But if easy takes us input, a Docker compose to describe your application, so UFC actually depends on or is built on top of kubernetes, but um for anybody who's experienced with kubernetes nodes, these really long complicated gambles can be kind of cumbersome for especially for a development team who's. You know, maybe not familiar with it, so uffizi abstracts, kubernetes API and presents it as a simplified Docker compose interface. So in other words, you don't need to know anything about kubernetes.

C

As long as you know, Docker and Docker compose you can describe your application and you physically will translate that or map it onto a kubernetes specification. So, there's a lot of abstraction and Magic happening behind the scenes that you don't have to worry about, um and so that's what this stuff is doing. It's taking these images and it's um uh loading them into a compose, a Docker compose file and here's the here's. The steps that you feeds is really kicking off here.

C

um These are this is a reusable workflow and if you go to the um the GitHub Marketplace you can, you can see this published as an official. uh You know the Visa verified Creator. One is what you're looking for, but this is this is this is using a series of things that we've already written out for the end user, and so all you have to do is copy. This example that's shown in here um and really just copy and paste it into your into your own repository and that's what we're looking at at right here.

C

That's what this step is, um but essentially what it's doing is it's calling this reusable workflow and it is passing this Docker compose file that we're generating it's logging into you, fizzy with a series of credentials and then uh yeah and then and what's happening in the back end. Is that you see the GitHub actions actually spinning up a runner. It's installing the GitHub, the uvz CLI, that uvccli is and using these credentials to log into uvz and uh deploy that um that application from from GitHub.

C

uh So so that's that's the that's how the pipeline is set up, and then I mentioned earlier, the um the docker compose file and so there's a Docker closed template that I have in this repo, and it has all those Services defined. So I've got my my redis service and my postgres service and all these different components, um and, as you can see here, these there's this syntax and and so that this is a variable, substitution and so what's happening in the background.

C

Is uh that get of action is building those images and then substituting them into this file? um And then this file is getting sent to you physi, so that um we can pull those images and deploy them for you um yeah. So so that's those are the two components. This is a Docker compose file and there's GitHub actions, and so let's go to our pull request here and we can see that all those jobs finished running and uh and the the.

B

A

C

Is that here's our URL and.

B

C

Got a preview deployed here at this URL that we can then go check out and we've got it password protected. uh So if you want to set that up for your um for your previews to make it extra secure.

B

C

Exactly I mean this is really high tech and uh State Secrets here and.

A

Do you set up the authentication within the app or within the GitHub action yeah.

C

So UFC supports URL, uh usernames and passwords and um actually I'll show you that real quick and it's right um right down here. You can see.

A

C

Username and password- and so um this can all be set within here, yeah great question and um okay. So, as we mentioned earlier, this is a this. Is a double dogs feature Branch I'm working on, so we want to test this works so I'm going to go here to the votes.

C

As you can see down here, we've got 15 votes, and if this works correctly, it should say 17 votes when we, when we vote so we're going to vote for dogs and we'll come back over here and we'll see, look there we go, we've got our 17 votes now, we've now we've rigged the election and hopefully dogs are gonna, are gonna win here.

B

um And then can you vote for cats and we'll watch.

C

You coming I can't I can't vote for.

B

C

You can change my.

A

Vote, I'm, I'm team cats just.

C

C

So um and what's cool about this too.

A

C

You can have different versions of this app running so, as you can see here, I've got two different branches deployed. So what's nice about this, is that you're not affecting your teammates work, so I can deploy on on open a pull requests check. My feature make changes in this case I'm, actually like storing data like in in the database, and it's not affecting my teammates deployment, because they're in isolated parallel environments. So this really gives you a lot of flexibility to test your feature and only your feature in isolation.

B

C

Really having any collision with other people on your team, so instead of having like a shared static environment where any change you're pushing is affecting everyone else's, you know testing. This is just your environment, for your for your branch and um and so it's a lot cleaner, simpler, more isolated and and just kind of a better way of testing.

B

It's awesome, yeah and I really love that there's. Finally, an open source solution for this I know in the past, I've used like Heroku um to try this and there's actually a question in the chat from subliminal. Is it okay? If I ask the chat, questions? Roselle? Yes, yeah, hey wondering what are other tools like uh it sounds like it's, a lovely Improvement for Dev experience and they it should have competitors.

C

Yeah, so uh a lot of what we're doing has been inspired by a lot of front-end preview applications, so you may have heard of netlify versus you know. Purcell. uh These are static. Cider, front-end, previews UVC is really a complement to those front and only preview sites, and so you can use it.

C

um You can either preview your full application on up easy or you can even have your front ends being deployed on, let's say, navify and then your API being deployed on euphz and actually connect those two, and so every time you open a pull request. You can have you know two different deployments going and being able to check it. So it's a nice preview tool that complements um some of the front-end uh preview application Frameworks that are out there. So that's um that's where I that's, where we've seen it use a lot as well.

A

Check if the chat had any more questions- or at least I wanted to say it out loud I did write it in the chat, but just in case, if people had more questions feel free to type it in the chat. um We're happy to like amplify.

C

A

And let Grayson know.

C

Yeah sure happy to answer more questions um and as we're kind of waiting, maybe on some questions the you know we can do. We can do some other things in here, like as I mentioned, we're constantly looking for um changes that get made to this branch, and so let's say I want to go in here and say: I want to change this color to, like you know, I, don't know, maybe like black or something right and yeah so I'm going to make that change as well and so oops say.

B

C

A

Oh can I, is it going to immediately apply those changes or is the action gonna run and then it's going to apply the changes.

C

I'm assuming yeah yeah, so the actions- that's a that's a great question, so the actions are going to run again. So here I am I'm pushing those changes up now this is a front end change, so I made a back-end change earlier, I'm, making an API change earlier now, making a front-end change um and if I go in here to these actions, you can see here that uh this this job has been queued and all those build push steps are happening again. So you, you asked the question earlier about synchronized.

C

What would the synchronized keyword mean? This is this is what it's what it's referring to? So it's now it's going to synchronize these changes. um It's going to kick off another series of build and push steps. Epc is going to continue to watch for these events again. These are event triggered.

C

um If he's watching for these events and then and they're going to deploy that change, and here in a second you'll, see you'll see a color change.

A

um And this one was fast. Third, third JP: oh, go ahead.

C

No, no, please go ahead.

A

I was gonna, say um uh someone who's watching named third JP was saying: how can you see container logs or the container logs.

C

Yeah, that's a great question, so um euphz uh has two options for uh to use it. The first option is a managed API service. This is a SAS tool that we offer at utc.com. This is free for individuals and small teams. We also have alternately. If you have your own kubernetes cluster, you can install uvz on your own infrastructure with a Helm chart. I'll show you the SAS version, because that's a little bit easier to visualize um here is our uh a dashboard. We have a series of projects. uh These are different.

C

You know repos or apps that I'm working on I can drill in here and um look at look at some of the logs here. So here's one from today from this is what we just deployed earlier. You can see all of my containers in here. These are all those Services. We talked about you'll see event logs. These are kubernetes event logs. So if there's any ever any issue pulling containers or something that'll show up here, and then we can drill down into these containers themselves.

C

As you can see, it's redeploying, uh if you check, see this uh remote workflow. So so it's previewing here you can see the step here, and so that means it's talking to you fizzy. You fizzy has just gotten a notification that these Services need to be redeployed, and so that's what's happening. Ups is already almost finished, uh just finalized.

C

The last one uh so I can drill down here into these different container services and I'll be able to see the um containers piped in here probably not seen him right now, because it's in the middle of a deployment, but as soon as that gets as soon as it's finished. All these all these uh logs will be dumped dumped into here and you'll, be able to see those so for troubleshooting. There's a container is doing something funny um you can. You can go and log into the UPS UI and check those out.

A

I appreciate how clean the UI is I feel like when I use like cloud provider or devops stuff like I, don't know what what it wants for me too.

C

Too many options, yeah options uh yeah, so so there we go so we just deployed it and let's see if it worked and there we go, there's our black there's, our black color and again. This has not affected my teammates uh environment, so I'm sitting here able to test a feature in total isolation, continue to push changes and uh until it's ready so so let's say it is ready.

C

um So let's go here so we've gotten our pull request. As you can see, we've got that that URL and by the way we gave it a thumbs up. So the initial deployment gets this rocket ship. Every time we push a new commit. uh We we increment the thumbs up counter here, so you'll be able to see how many times it's been updated.

C

uh So let's say we like this and we say hey. It looks good. Let's go ahead and merge this. um Let's go and confirm the merge that is similarly going to kick off my Pipeline and it's now going to see hey uh this. This branch has been closed, so it that it merged into the to the main branch and um and then you fizzy, will turn around and delete, delete this environment, and so this is what really makes it nice about you?

C

Don't just have this mushroom farm, like environments being stood up everywhere, that never get cleaned up? This close pull request. Event is going to clean those up, so you don't have to worry about those. You know: orphaning, a bunch of environments.

A

So if I clicked on that link, it would then there would be nothing on there anymore, yeah.

C

So if you can see here, these are those actions that are getting kicked off again.

B

C

If I, if I go back to this, that to that pull request, issue and actually just uh pass it here um here in a minute once once the pipeline finishes, if I hit refresh it's going to give me a you know a a 404 and it's not gonna, be able to find that. So it's going to be torn down, that's awesome, yeah and so, and these are all secure.

C

These are https, so they're UPC is handling all the networking, the SSL termination, um and so this can be used for, for um you know, applications much more advanced than like a toy. You know a toy app like this.

A

That's awesome and there's another question.

A

um Someone said: can I use ufizi for other types of applications, for example, say a CLI tool where by you fizzy builds my co live for each pull request and I can download and test the CLI binary.

C

So yukisi is best suited for web applications, the things that are container to be able to be containerized in web applications. You can it's probably best suited in your situation, for maybe an API that you want to be testing against making changes and then using a CLI to to. You know hit that endpoint that URL, that that um uh that we make available in terms of building and testing a CLI itself. It would be something that's if it's something that can be containerized and it makes sense to deploy in the context of a web environment.

C

You could I, don't know exactly what your use case is, but those are kind of the the guidelines that we recommend and if you check out our our main application, repo Utz app. There are examples in there there's.

A

C

Examples and links to further reading um as well on our on our repo there's links to um a lot of things like our documentation. For example, uh you go check out and read some more about that here.

B

I guess, since we're looking at the documentation now, how can people contribute.

C

Yeah, so if you, if you go to our main readme there's a series of community links here, um slack channel is a great way to just say: hi pop in ask questions. um Ask hey: do you support this? Do you? uh How would I do this or that the other? We also have a public roadmap. You can check out if you fizzy, doesn't support any something that you want right now. Please, please open an issue and we'd love to hear from you.

C

um You can check out uh what we're working on what's coming, and so those are all great ways to kind of get a pulse of the project and then um and then all of our all of our open issues are also available here uh at uh we have we have a couple projects here. This is our working project board, so these are like open issues we develop completely and open. uh You can see what everybody on the team is working on, lead comments.

C

um If you have suggestions and uh also there's a whole there's a bunch of issues and ready for development, so we try to use some tags to kind of um organize What's. um What's what's needed to be done? There's good first issues.

C

um You know where our application is written in Ruby and golang, primarily um but uh yeah. So if you're interested- and you want to contribute, please please check out those resources.

A

Awesome that was great. That was great. uh Let me just check uh somebody asked this I think this is more geared towards me um and I'm. Just gonna say it out loud for anyone who's not watching the chat they said, will they be able to re-watch on YouTube?

A

The answer is yes, I'll probably put it up in like two weeks or so, and also hi to other folks like Lazar, thanks for watching um and my other question to you, which is, we talked about this a little bit on the Twitter space, but would love for you to share here as well like what was your motivation for building ufizi.

C

Yeah, so it really started out um about we. We really wanted a way to be able to, as we mentioned preview things before they get merged, and we started talking to a lot of folks and we found that this was a bigger problem, especially related to Shared test environments. Typically, you know when a team is all using a shared dab or test environment.

C

You've got a lot of commits from different developers, and sometimes you have to wait to get your code deployed to a cloud environment, and so that may be a bit of a bottleneck right there. The other bottlenecks that occur are you've, got a bunch of developers they're pushing commits. Inevitably one of those commits is going to break your QA environment, your test environment.

C

It happens to every team and it's a huge slowdown, because then not everyone else can't test their feature properly and so they're all blocked, and then typically, your senior developer is the one who has to go in there infrared what went wrong and then try to revert those changes and that's a lot of time wasted.

C

For you know one of your most valuable employees or your teammates, and so we really saw a need for or I as we mentioned, isolated parallel environments that everybody can have not just their own environment but like every feature, Branch can have its own environment and we started talking to teams and and a lot of people had recognized this problem and had set out to build a solution, um and but inevitably everyone we talked to was not happy with what they had built. It was like yeah, it doesn't work great or it's.

C

You know a little buggy or whatever, and so that was a strong signal for us that this there needs to be an open source standard for this. So teams don't have to keep Reinventing the wheel building this from scratch, um and so that's what we set out to do, and so earlier this year we open source this and really have been trying to get a lot of a lot of other people who who, um like we mentioned people who have thought about building this or have have built it contribute.

C

You know your best practices, your designs and uh try to create. Like a standard for everyone, so that everyone out there can get like an environment to the service capability.

A

B

A

B

So many more questions jumping in here everyone's very curious about this person um yeah this one from is the pipeline code itself, subject to Source control.

C

So the question is, at least in the case of GitHub actions and by the way up easy supports, not just GitHub actions but gitlab CI, Circle CI, any CI provider. It's designed to work with, and in the case of the you can have actions all of it is version controlled in this directory. So if, if I want to make changes to my pipeline, it's all being committed here tracked here and all those changes and so um and every time I, you know open a future branch.

C

Of course, I'm copying this this um this workflow and and all that's being tracked with kids. So that's! What's nice we use git as a source of Truth uh for everything, and so it uh if you're, making changes to your pipeline. It should all be tracked and get yep.

B

Thanks for the question please.

B

And then can this be used privately in-house with bitbucket.

C

So we in the case of yes, the answer is yes, um it depends on where you're pushing your your container images, but um if you want you can set up these easy to pull uh from con from your registry. Let's say: you're: pushing images to Amazon ECR, um if you're, if you're, if you can, if you can add a uffizi action or job or whatever I, don't know what you know: CI provider you're using um you've easy publishes. Our CLI is a Docker Hub image.

C

That means that you can take our container image of our CLI and wrap it in your provider of choice. So in this case we've written a GitHub action around our CI um same thing: we've done in git lab in circle CI and all these other CI providers generally have the same functionality. So you can just take our open source, CLI and wrap it with your CI provider and and then you can deploy from from any uh any provider. You want from any container registry you want. So, yes, it should work.

C

It's not it's not just with GitHub or GitHub actions.

A

I love how flexible it is. Oh Abby was going to say something.

B

No, the same thing: yeah.

C

Not a spread competition like the the git providers, but we do. We do try to be as uh agnostic as possible. Yeah.

A

I'm, okay, I, like that, you know, because not everyone's on GitHub or not everyone's using GitHub action. It's like that's just the reality, so I.

C

Like that absolutely yeah.

A

All right, everyone said Thank, you like blaze, said Thank, you and then cruising Carmel said. Thank you as well. um I feel like it's a good time to pivot, like keep the questions coming, but to Pivot from the technical Parts I learned a lot to like the more non-technical questions that I usually ask yeah.

A

um My first question to you is: what's your favorite food.

C

Favorite food well I'm a I'm, a southern boy, so I'd probably have to go with some barbecue, maybe a little cornbread and some some sides with too much cheese on them. I guess is probably my it's probably my choice.

A

B

A

Wait: Abby I know you didn't prepare, but I'm also curious.

B

A

B

Well, I kind of want to eat the second half of my sandwich that I have sitting here.

B

A

B

As of like this minute, that is my favorite design.

A

I feel like I, can't wait to go back to finish eating yeah.

B

No, this was an excellent break. Thank you.

A

All right and then what's your favorite, oh wait. I always tell everybody my favorite food, so the the chat doesn't need to know they're, probably tired of it. um But what's your favorite Beyonce song, you said you're keeping a surprise for me. Yeah.

C

I told you I told you I, you asked me this earlier. It's like be right. I was like Beyonce I. My Destiny's Child is music of my year my year, so I had to go back to that. It's either it's either gonna be Survivor or maybe maybe bills, bills bills. You know I.

A

B

C

Might be a scrub today if I wouldn't.

A

C

A

Are good I, love, Survivor? Actually, no one's come on the show to say Survivor. Usually people are like they like bills, bills, bills, yeah,.

C

A

C

A

That's a good song that.

C

A

um And Blaze asked another question he said: uh so: will the docker image run on arm processes.

C

uh Arm processors, um processors, sorry so Raspberry Pi easy is deploying things onto a kubernetes cluster and for the managed service version this is being deployed onto it's. Actually Google contain uh kubernetes engine running uh Linux as the OS.

C

um That's a good question, uh so if it can run on kubernetes and Linux in a container, uh the answer is: yes, sorry, I, don't know if that's helpful enough for you, but but that's what the.

B

C

Option is, of course, to install it on your own uh infrastructure and there may be some configuration you could do there, but I I can't say for sure.

C

It's not something. We've tested put it that way.

A

Yeah I was gonna, say blaze, you could try it and then you.

C

Know open an issue and then we got it happen. Yeah.

B

Build your own Raspberry Pi cluster, see if you can deploy.

C

That right, yeah I've heard of that.

A

Also Pedro Pedro, he came on uh maybe last week or the week before that he said. Please don't be like me and Confused Beyonce with Rihanna.

A

Pedro's awesome um and then also hi will just saying hi to people. um uh There looks like it looks like there's another question or somebody saying something: okay, let's see Sandra TV said since it's a one-time setup, try to make your CI task so powerful, like you're, not limited by the number of workflows. You can have check linting check form. Oh, so it's not a question they're, just making a comment: cool.

B

C

B

Even more powerful yeah.

C

Yeah there you go there, you go yeah, so all those examples are in there. If you're, if you're new, to get up action, then we uh we're actually working on getting some documentation put out for some other providers here soon. So uh please give us a follow: give us a star and or join our newsletter either one you can find out about updates um we're constantly like shipping features. So even if we're not supporting your exact configuration or if you're waiting on some docs, um you can get notified when those windows get released.

A

Yep and I put the URL on the oh. Someone has.

B

A comment well blade says that kubernetes runs on the Raspberry Pi, no problem, so I guess yeah, I, guess: okay,.

A

These are really cool. I just bought a Raspberry Pi, so I want to try that sorry go ahead. No.

C

That's I was just thinking, maybe set up your own data center of little Raspberry Pi they'll, sit around and run and run in kubernetes on them. So that sounds like a fun fun piece of what somebody listening out. There wants to do that and share come.

B

B

Project for someone who's not me.

A

I there's another question: I, don't know if it's like really relevant to you fizzy, it's more like software engineering, workflow I, think, but they are asking like when to release a tag.

C

Yeah, so like not every team, does you know uh tag releases, uh it's a way to be able to kind of version. Your your releases um generally, it's it's when, like a some, some teams tie it to their Sprints or their ethics um process or when you're, releasing like a significant version, update.

C

um There's a lot of literature out there. That can probably explain it better than me, but uh I assume you're asking, maybe because we have releases uh noted that are repost.

A

C

This is a good.

B

A

Will say this is a definitely a tough question and I'll say like, like you said, go ahead and like go read about how how other, like engineering teams are doing this and then um figure out the.

B

Experiment, if it works for you, yeah versioning might be something to look up, um but yeah tags are nice, though generally yeah.

A

um Cool all right so go ahead and start you fizzy, y'all or, and if you want to, if this project sounded so cool to you, go ahead and like contribute or try out the the GitHub action or any of their tooling as well, um and then also go ahead and follow Grayson on Twitter at Grayson Atkins.

A

um Anything else that you would like to plug Grayson, oh and there's more questions. People people are really interested hold on before we plug um someone said, can I access containers directly in my ufizi environment.

C

So that that feature is coming soon, we have on our roadmap um SSH your shell access to the containers, um so you can remotely log into them uh right now. It's read only in the web UI, which is the container logs, but here pretty soon you'll be able to uh log directly into them.

A

Nice, nice and Blaze pointed out something pretty good, I think for Sandra and everybody else. Who is wondering about cicd, um they said, or he said, go ahead and read up on the humble Farley book. Apparently that has good information for for beginners to see ICD all right. um Do you have any other things you want to plug things that people should look out for when it comes to you or you, feasy.

C

No I mean next week: we've got some some testing um uh videos coming out kind of showing how you can get you easy set up with your test environments.

C

um We got some more stuff coming up with illustrating how you connect, netlify or sell preview, front-end previews with upz API previews, and have those talk back and forth and we're doing some exciting stuff here, pretty soon, showing like an in-n testing using Frameworks like playwright or Postman or Cyprus, to be able to stand up.

C

You know multiple PR environments and run your Indian tests shut everything down afterwards, so really trying to give people playbooks for how to really just supercharge their development and testing efforts so be be looking out for that and again that those are all be announced in our newsletter and we'll put some blogs up about those too.

B

Nice perfect, maybe a good thing to close this off on is the question that I was very curious about yesterday, but why the name of fisi.

C

So uffizi is inspired by the ufitsi art gallery in Florence Italy. How is this some of the great works of the Renaissance? And uh you know just like they're uh showcasing these great works. uh You know you feed these away for teams to Showcase their their uh their application, their works of art in the digital world. So that's kind of the uh it's a little bit of a holdover from another project that we that we were working on and um we just we just like the name just because it's fun to say so.

C

It's easy, like you, fizzy.

B

Hello I just wanted to hear you say that again.

A

That's that I think that's dope, but I don't know.

A

All right that this was great y'all I learned a lot. I think the chat learned a lot too um so I really appreciate you hopping on thanks to you also Abby, for for coming through and hanging out with us bye y'all thanks for watching right.