Red Hat OpenShift OpenShift Coffee Break | Red Hat Livestreaming, 13 Apr 2022

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: OpenShift Coffee Break: DevOps with OpenShift

Description

Get your espresso ready for the OpenShift Coffee Break as we talk about everything DevOps with OpenShift! Our special guest is Wanja Pernath, Solution Architect at Red Hat, and author of the new "Getting GitOps" e-book from Red Hat Developer. Join the discussion to share architectures on OpenShift with Tekton, Argo CD and all Kubernetes automation!

Twitch: https://red.ht/twitch

A

Here we are welcome everybody good morning. Welcome back to the openshift tv coffee break from this session. Today we have a special guest, but before introducing it, let me introduce all the stuff here. My name is natalie vinto, I'm product marketing manager here at openshift and hey jafar andrea. How are you good morning.

B

A

Finally, thank you, my wonderful co-host here on openshifttv and we have a super special guest uh book author. Also we're gonna talk about this awesome book uh that it was out from a few days. Welcome one yeah, hello, welcome and so first thing, vania vanya is the correct pronunciation right. Yes, yeah! That's right! Okay! uh Mario, do you want to introduce yourself yeah.

C

Of course, um so um I'm yeah I'm working as an partner enablement manager for technical stuff on openshift, from the development perspective at red hat.

A

Awesome and uh the reason why we invited you mania is not only because you are cool and you do lots of stuff a lot. We know you do lots of cool demos, look workshops, but also because you are out of a book and I think andrea has a copy.

D

There copy of the book getting here.

C

Very important the monkey in there very important.

A

So for the monkey- and I I would like to to say a lot to the first, I would like to say hello to the to the chat. Welcome everybody. If you have any question, uh please stand in the chat we will answer in the show. The topic of today is devops with openshift and we talk- and we take this opportunity to talk about also the the book that one uh vania wrote and let me also share in the chat this link, so the people can download for free right vania.

A

This is a free ebook that we'll talk about. I understand talk about gear ups. Do you want to do a little introduction on the context.

C

I would like to so um yeah. It is about understanding and using git ups, but it's not um theoretical.

C

So for from my perspective, it was very important to have something like an a practical guide, a practical blueprint, so to say um to your guides to githubs, which means it's discussing um a well complete use case with a rest service um based on on quakers at the beginning and then, of course, the rest service is using a database and then you're going through um first of all to the through the development of that rest service with quakers and then um down to the complete road until you are using tecton to do um your development pipeline staging pipeline and then at the end, setting up argo cd in order to make sure that you have something which is automatically deploying the complete environment which you're using.

D

How was it, how was it to write a book.

C

Well, um at the beginning, I didn't want to write a book. This is quite interesting because um the first thing I wanted to do is because I do all this in my day-to-day job when I'm enabling partners for redshirt from technical perspective- and the only thing I wanted to do at the beginning was just taking notes. So I'm talking a lot when I'm doing my enablement sessions, and so I I needed something um just yeah well to have my notes to prepare myself, for example, at the beginning.

C

Well, and then I thought well um after after a session, I had so many questions coming from from the audience I'm asking about.

C

Well, I'm I'm overwhelmed with the news of um open shift and kubernetes, and then there are so many tools like text and like um argo cd, it's way too much, so I thought well, it might make sense to start writing a blog entry about this, just one, a small one yeah, so where I'm able to point my attendees after that, then to um to that blog entry, so that they are able to read it right and well, then um it was uh two block entries about this.

C

Then I decided to have um three because I also wanted to talk about um customized and things like this at hem charts, because this is also part of the cicd chain, and then I thought well, if I'm doing all this, I also have to talk about um text and pipelines. So I also wrote a complete um article on this and um between the third or fourth and the fifth chapter or blog entry. My management came to me asking me well, why are you not writing a book out of this stupid? They said.

A

C

Yeah well, and then I um got in touch with um other people inside reddit who were just um writing a book. I think you were one of them natalie um and um so well yeah, so I asked them and they were so friendly.

C

I wrote professionally with them, with with the with the editors um at red hat, it was so nice and I introduced githubs with them so at the beginning. um So I did all my writing with um markdown, of course, because it's a technical book, so I can write markdown and um the initial thing was to generate a word document out of it to send it to them, and then I got it back which was completely red then, because what mistakes I did and so on, and then I decided well. No, this can't be.

C

I do don't want to um to to export imports and stuff like this, so I ask them if they understand to use git and they said yeah. Of course I I know gitz because he's a technical writer as well um the editor, and then we decide to use uh the complete kit. So if you want to have a look at the git repository of the complete book with all the sources, you can find it publicly available on github as well on, under my account.

A

Oh wow, so uh you did.

D

C

It was interesting I can tell you.

B

So so did they ship the real copies with the with pipelines also or how do.

C

No, unfortunately, not this is just by well um using git as the central repository to to write all the chapters and so on and to get the change back.

B

Yeah, I think it could be uh good if we at some point share the link.

A

Yeah, it's already shared and look look. This is the page and uh one I was already sharing the in the the screen yeah.

A

We put the link on the uh all the book, but now I think uh we have also this yeah.

B

I mean I mean for the.

C

Github for the github- yes, so this one! Let me I shared here. I can't access the.

A

B

We can share.

A

B

A

You very cool, so that's the source code of the book exactly yeah also.

D

So the process itself was an interesting one in initiating.

C

D

And working with the other technical writers and editors collaboratively.

A

Using it very nice, very cool, very cool- and you mentioned uh if you can go back to the cover of the book, uh so this is the landing page. We put the link in the chat you can download it for free, it's an e-book, so you get a pdf just register for free at developer.com, which is the editor, the publisher and uh look at this cover. You say the the monkey.

A

This the sign is from our awesome colleague colleague, shout out to kalik we made very nice graphic and, uh and also the content is very cool. But vane. Do you want to talk about the contents? The title of the show today was devops with openshift and techton and argo cd. uh Can you explain us? How do you compose your devops with openshift with those two open source projects.

C

uh Yeah well um so it was. uh The discussion was quite nice with um with all the internal people and um understanding getting the title um of the of the book was also quite interesting, so I at the beginning, I wanted to have something like um understanding, githubs or githubs from scratch, or something like this, but this was not necessarily what um so. There were. Other voices who thought getting.

C

Githubs is way better for this, um and one thing which is quite interesting is this monkey in the title in the title which you can see there, because um I wanted to have a head scratching monkey sitting in front of a jungle where you have um signs saying well, this is tech, 10 go there. This is agua city go there. This is pokemon, go there it's on and then that you're having I'm going to have a monkey which is like this okay. Where should I go now right and um at the beginning?

C

uh So we had a discussion so to say on how to um create the title, but, uh as you mentioned already, he created the monkey in there and I'm so thankful for this, because this is really important for me.

A

Nice, nice and I'm looking at you know these books cover. Can. Can you show the list of the? What are the topics of the book um I'm reading? For instance, you know uh beside basic kubernetes, then there's a a bit of opposition. Then we have, you know customized tecton, argo cd helm. So there are lots of uh cool tools right. Yes, they.

C

Are because it's um in nature, when you are when you would like to set up a new project or would like to reuse an existing project, um then well, you will be um thinking about how to um well, first of all how to do ci cd with a complete with your new new project. So um would you like to still use jenkins? Would you like to use tektin for this? Would you like to use whatever anything else out there?

C

um So I I was thinking about answering some questions and doing giving in um a demo on how to use this. Naturally, then, um also when you're thinking about um distribution of your applications, for example, if you're in, um if you're selling your app to your customers or something like this.

C

So how are you able to distribute or to sell your application at the end, then you, of course you need something like helm or operators, or something like this just to make sure that your customers able to download and install the the application which you've provided to them.

C

So I'm also discussing the differences between helm, charts and operators and when to use which and what and of course, every chapter of the book. Every topic has also um a practical um example of this, and all those examples by the way can also be found publicly here on github.

C

So um just have a look at the book example repository where you can find more or less everything we are starting with. um So when you are new with openshift and you are using for, for example, the source to image um environment in there to build your application inside openshift. um So then the next question is okay. Now I have my development environment.

C

So how can I go over to the test environment so that I can um stage my my development, for example, um for this I was thinking about well um either we are just exporting um the deployment, the image stream and blah blah blah blah blah, or we are I'm going to use um openshift templates for this, so I'm discussing also templates and also thinking about when to use templates. So if you're in a mixed environment, you can't use um templates because um openshift templates are unfortunately auto shift specific.

C

So you can't just use um object templates for this, so there must be something else, and this was then customized. Yeah customized, in my opinion, is one of the core technologies when you are thinking about ci cd with um in that kubernetes environment.

C

So I have a complete chapter of um customization files and customization things in order to well make sure that you're able to deploy your application from scratch at the end and then, of course, then the distribution thing. ah Thank you exactly. Then. The distribution stuff is um well. If I would like to make my application publicly available, then I need helm. I need kubernetes operators and things.

C

Then, of course, I'm digging digging into cicd using attacking for this, and the final part is then well if I already have everything if I'm using customize, if I'm having all my files or the um metadata files, which I have for my application already available in the git repository.

C

So wouldn't it make sense to have a tool in between my git repository on the one side and um the kubernetes shift cluster on the other side, and this is then naturally, of course, an argo cd which is um watching on one side and is deploying all the changes back to um back to kubernetes overshift in this case.

C

So this is all the way down so to say.

D

So, in a sense, you analyzed various options, but you also in the end provided a recipe starting from scratch to do devops. Sorry uh from the beginning, to the end. Obviously, you took a example of an application. You chose quarkx, but uh that's it's applicable to many other to many other cases.

C

Exactly yeah, so um the book example repository could easily be used as a blueprint for this. Of course, if you have a larger application landscape, not just one microservice, which is connected to a prosperous database or so um then you need to think differently from time to time, but for each of those microservices which have one or many um relationships to to external tools like a postgres database like sso or key cloak whatever, then you could easily use this as a blueprint.

C

You can go through the complete example and at the end you have an um yeah. You have pipelines, you have the understanding, the basic understanding of using um argo cd with this, and you are able to easily deploy your complete environment.

C

A

That's pretty cool, that's pretty cool and I'm sure you know um I'm sure we, the audience, also would like to see some of those examples. But before that uh uh vani have a question um I see you mentioned that also openshift templates um in the book. Do you think that the templates can be uh partially at least substituted by elm charts or do you think are still relevant from the devops story, at least in openshift.

C

So from an open shift perspective, I personally would use templates not for doing ci, cd or pipelining things. I personally would use openshift templates for um setting up um samples for my junior developers, for example. So when you are um so when you, when you are going to create a new application for example, then most of the steps are always the same.

C

So you need to um to find the baseline for something you need to connect it to a database or to an sso, thingy or whatever, and all that can easily be created as an as an as an openshift template. So, like all, the examples which are available inside openshift by default um can also be used as an um from an uh from an uh template perspective. So the nice thing is templates are integrated into the ui of openshift, so which means any developer is able to easily use that template.

C

um There's a description. There's an icon. There is a list of properties which you need to um to to define or to change of course, um and then you can easily start from scratch a new um application, so um I would say from if you're an architect or a senior developer, and you would like to create some um some templates for your junior developers. I would use this one- those ones: okay, okay,.

A

Yeah, that makes sense it's uh still valid. You know to create some resources from from the from the web console web console that I know from jafar. It's very customizable right, jafar you're, working on a set of plugins in the web console.

A

Do you want to say something about it.

A

We cannot hear you.

B

A

D

uh I wasn't yeah.

B

So um yeah, that's um actually a a very cool feature that we added uh recently uh on uh the openshift 410. uh It's something that we we wanted to have for for a while, which is the ability to add your own pages to the openshift console. So you can provide a very customized experience and also maybe bring data from other tools into the openshift console, uh so I'm actually going to have uh a show on it. uh This afternoon, oh with the ask an admin, uh it's on the level of hour.

D

This afternoon,.

B

At 3 pm cet and it's basically um yeah, we call that the console, dynamic plugins and it allows you to write your own plugin in in react. So we do provide all the framework and you can then have your own uh things appear in in the console, so yeah. So, for example, what you mentioned about the templates, uh you could also create your own experience like if you have a developer workflow that you want to have uh for your for your you know as a routine to onboard new developers.

B

You can also create like a three steps or four steps: wizard that will be really customized for your developers and it can use under the hood, uh whatever you want, like templates or or hem charts, or something like that and that's actually one of the plugins that I'll be working on is basically how to have a yeah, a simple uh steps: wizard, but something very customized.

B

uh Depending on on what you wanna wanna do so yeah. I do agree. The templates are still something to kick off uh the the experience with the people that are not very familiar with the platform, whereas with hem shots, you can do a little bit more sophisticated things and they are probably a bit more adapted to a ci cd pipeline because of everything you can do with injecting the values and having the customized things, etc.

A

Right more dynamic wires, where templates are more static, yeah that makes sense.

B

They offer a little bit more flexibility. I would say.

A

That cool it looks uh andrea. It looks a very cool day for openshift tv. We have vania today talking about the getting hitops, then we have uh jafar at the level of power doing this live demo of customizable web consoles, so stay tuned on openshift tv, but still vania. Today, our today, we are here for devops with openshift, uh and can you please show us the the code as as we say here can? Can you show? Can you do some live demos to us? Please, like we would like to see that in action.

C

Yes, of course, um so if you would like to share the screen, thank you so much. um So what you can see here right now is um nothing so all the um default projects, um so it's not has nothing to do with what I'm going to show you right now. So I need to provision um my demo environment now and by the way this openshift um is running on my single node shift cluster running on an in german, you would say brule versus so it's called uh very how.

A

Do you see that.

A

Can you write it? We.

C

A

Written down and we're sharing the chat just seconds.

C

Let me go to the chat. um It's called.

C

Because this is very small into nuc um environment and it's quite loud but um yeah, it's it's nice, so it has 48 gig of ram and something like this. So it's it's really nice um but anyway. So that's my open shift single note uh running on um so now, let's install um my ci environment first.

C

um So let me go to this one, so this is coming from my um example. So just download or clone the book example repository. And then you see here in the um githubs repository you can see argo cd and tactile.

C

So one thing I'm doing first is: I would like to install my tekten environment. So I do an oc apply minus k because those folders are customized.

C

So for this, so now all the tasks I'm using are being installed or the book ci environment is being installed, persistent volumes and blah blah blah blah blah everything is being installed now um and then you can see it basically here.

C

So if you go to the book ci application now you can see first of all, the nexus environment, of course, is being installed as well, because it's a java application and when you're building it um on openshift, you need to well find a way to store, not the complete internet but parts of it right based.

B

C

Of the internet half of the year yeah.

D

C

So now, of course, we have two pipelines being installed here and the death pipeline. The stage pipeline um just have a look at how those two are looking like. um So the dev pipeline is cloning, the source code. So the book example repository um cloud conflict. Is the person service config repository which being deployed to is the git ops repository so to say um where this one is very important, um so we have everything what we need in order to install our um our application on a new namespace is part is stored here.

C

um So this is one thing and then, of course, um we are packaging, the application we are building the image. We are pushing the image to a quay um registry- query io. um In this case it could be any um any registry, of course, but um I'm using public way because um well I use it, I'm used to it and then we are extracting the digits um coming from the build of the image and we are updating the um git config repository with that image, digit yeah.

C

So when we are going back to um person service config, um we are changing the overlay um depth environment. So this is um the difference between the base and the depth environment, for example, in this case here in the in the customization file, you will see the um the new image digits of the current build you're going to see here.

A

So this is the point uh sorry to interrupt. That's the point of contact uh uh from the ci world with tecton to the cd world, with argo cd right exactly exactly yeah.

C

B

Yeah, that's all good yeah. Sorry yeah go ahead, but I'll have a question uh just after afterwards. Oh.

C

Okay, so the difference also, we are used to have an um well a pipeline which is doing everything, so we have a pipeline which is downloading everything which is building everything and then which is deploying everything at the end. This is um the way how we are used to do. Ci cd at the end and um githubs is a little bit different by github's.

C

You have to think about who is responsible for deploying the actual application at the end, and in my opinion this is argo cd when we are thinking about this, because we have everything what we need is stored in git. So when argo cd is going over this, um it is reading. Everything from git is merging the the config and it's deploying it in a new name space.

C

So I'm not actively deploying anything with the tecton pipeline in this case, but I'm um just instructing it to update um the image digit in this case and then argo cd is taking care of everything else yeah. um So you had a question.

B

Yeah sure so, can you please go back to the the screen that you were just showing here and maybe show yeah? Can you show the last uh comment, change uh where it says updated the image digest: ah okay, yeah just so we uh so so we see exactly like what, because I think this is very important if you can't show like the last comment or okay. So that's basically, you are going to create um yeah the new type.

B

So is this based uh so do you have, for example, a track to the commit id or something like that, or is it just like the the latest tag of the image that has been created? This.

C

Is just the latest type of the image which has been created so when the build pipeline um is building the image and is using the image id or the image digist so to say um to update the customization yaml file in github right.

A

Talking about tacton, I guess vania. This is a tecton task that has a step where you basically extract from the result of the previous text on task, the hash of the container image, and then you use customize uh to to update right. Okay, exactly.

B

Yeah and so yeah thanks, so it's it might seem too detailed. But I think this is a key point, because that's actually what makes it work it's. uh It's like. I have something uh a sort of template and then, whenever you change that value in that specific repo, it's going to trigger the the deployment by arco, cd or what we call the the synchronization loop and something else. I don't know if you mentioned it, but you basically have two repos. You have one for the application source code.

D

B

You have another one for the the git config that is used by the githubs tool or argo 3d, and this is basically a common approach. Nowadays is basically, you separate the code? Yes, so so you don't have a pipeline that gets triggered whenever you you make like a code change. Unless you want to have something you are going to trigger the deployment uh I mean the deployment of the application gets triggered when you make the change in the config repo and not in the application, uh repo necessarily.

C

Exactly yeah, so um so right now, just the ci environment is um set up for this, so um it does nothing. So what I'm not doing um right now, but this is just because I have my um openshift cluster running inside my own network, so um I do not have a trigger so.

D

C

Yeah um I've tested this and it works when I'm using um not openshift inside my own network but somewhere on amazon or something like this, but for this environment and because I am living in an area where a good network is sometimes some kind of luck. So I'm trying to use everything internally. You.

D

C

um So this is the only thing I can't um show, but it works.

B

Okay, yeah I mean, but it's just the principle. Is that yeah you.

C

B

Commit to the application code it's going to trigger the the ci pipeline and then that same pipeline will update the git config repository, which will in turn trigger the pd to deploy the application to whatever environment. Exactly.

C

Yeah cool, okay, so one thing more, I need to do in order to set up the um um the tecton environment now or the ci environment. Is I've created the um bash script, which is also creating um some um service account?

C

So if tecton is so wet when tecton is starting to build something, it runs always in a in the security context of a service account um and because the build pipeline also makes use of um a docker push so writing to a docker repository craig in my case- um and it is also writing to the github repository- I need to provide somehow um all the um to the credentials to those two- the cigarettes yeah exactly yeah, and this is done by this by the way um they are two.

C

um There are two uh passwords in this, but I'm always frequently changing them. So you can copy paste this if you want but um yeah well, you shouldn't yeah.

B

C

I have to I think um so this script is um as mentioned. This is just um doing the same. What I did with oc apply minus k already, but it's also um creating the service account. It's creating two secrets. I want to access the um doc, the the docker repository, the cray io and one to um um create the secret for the um for github in this case now so now everything is in here. So the next thing I need to create is um the argo cd environment.

C

So if I go back to shift um having have having a look at the our city environment here, you can see there is nothing right now. So we are. This is argo cd, which is coming with um with the openshift githubs operator, which you can install um in there, but then, of course, you need to create applications which are then the mapping between the git repository and the kubernetes instance openshift in this case, so um I haven't created anything right now, but I do this right now. So, let's go back again.

C

I do an oc, apply minus k um on my rbcd folder. um So now it's creating two namespaces book, dev and book stage and it's creating two role bindings for the argo cd application reconciler. So the service account in there so that this service account is allowed to write and update book dev and book stage yeah and then, of course, it's creating the two argo cd applications book, dev and book stage. So let's have a look back here now those are created in in argo cd.

C

We have book dev which, as you can see here, some where our details are detailed.

B

So sorry, just uh this uh one question about the uh the the role binding. So so is it a role that you have created uh that allows like right access to the namespaces or or is it a default role that that you have.

C

There is a role, so the um argo cd um user, so the internal use of argo cd um that needs to have um added access to those two um namespaces, and this is what I did here. Okay now so now you can see that the book def application is okay, I'm already assumed successfully with my openshift environment.

C

And the same is also true with the book stage environment, so it is all already done and you can see the synchronization was done um and um there is also yeah well, there is no problem, as I can see. So, let's go back to open shift here, have a look at the topology view of the book def program namespace and you can see. We now have the person service.

C

Here we have a database which is by the way, based on the crunchy data, um prosperous operator, blah blah blah and both are connected to each other already. And if I'm, um if everything has worked correctly, you can see it now. um We should have the person service which already have content in the database, because augustini also has the um has a hooking mechanism.

C

So when you are deploying a new application, you can create a hook which, for example, is uh being called after the synchronization happens, and I used that after synchronization hook to fill the database with some um with some life yeah. So we have four entries in there for singers.

A

And since this is a live demo of anya, can you share the link in the chat, so we can share to the chat to your uh your url, so the people can see it. It's really live it's in your cluster, but it's really apple ryan. Yes,.

C

um Otherwise, it would work, of course, but um it's my internal cluster running okay yeah, so that doesn't work. Unfortunately,.

B

Yes, it's life. Well, this is just our point right.

A

No just kidding it's not powerpoint.

B

uh Yeah, so can you show uh a a bit more about the the hook that you are using because I think that's an interesting thing.

D

B

Only the first time or does it run whenever there's a thing so just to um it, runs.

C

Always every time there's a swing, it is going to run, okay, second, but um um just a second. That was the wrong one. This one sorry um so, but um it has um a check in there. So if you have a look at this here so see this is the um the post sync.

C

Okay, let's make it a little bit bigger? Yes, okay, that's better! um So.

A

C

Zinc um hook, which is just nothing more than a typical kubernetes job in this case, but it could be anything yeah. So when you're, when you want, for example, you could also use um a tactile pipeline which is being executed as a post, sync hook, yeah, so that would all work.

C

So what does this thing being doing? It's? um Well, it's taking the service url, which is the internal url, and it's first of all, it's checking if there are already some persons in the database. um If no, then we are going to fill some data into the database and if yes, then well, it's going to be ignored.

C

That's all here.

A

Nice, that's pretty cool and very useful vania. If I can, when you work with databases, if you think about the precinct uh now we're seeing the posting no but the precinct, you know it can be, some database schema upgrade or something like that, and the posting is something like you show that so it's very important to also put those kind of mechanisms like oops.

A

We we put in the chat the link to the documentation are go cd, documentation on hooks if you like to check out, but this is a real example on how it works on a real use case, right application, database and and also like available databases since you're using the operator right. It's kind of a I see a stateful set and there are multiple pods.

A

So it's a production rate like something uh use case. Somehow.

C

Yes, exactly um here, you see, you saw this is the um the personal service, it's really a real rest, crud service yeah. It has implemented, I think, um or I think I implemented all the methods which are necessary so creating a new person um updating that person deleting a person and so on. um So it's a real rest service and it is um connected to that database um over there. So if you are interested in how I'm provisioning the database, um this is this one um postgres yaml and yeah.

C

It is a it's coming from an operator which is the crunchy data operator, I'm using here, um and the nice thing is that I'm installing the database just as an um on a declarative way. As I install my application declaratively, I'm also installing the um the database decoratively- and this is quite nice yeah. So.

A

C

Is everything I need in order to deploy the database and the rest is being done um interactively behind the scenes by the um corresponding operator? There are several operators out there, but from my own youth, when I still was blonde and not gray, I'm used to do prosperous, stuff and oracle, of course, but possible stuff. So I was using a prosperous um for for that environment here and by the way. um So when you want to build the environment now, so um we have no personal service 1.6.0 whatever based on quakers, something.

C

So I'm now changing this a little bit. So let's go here: um let's go to the index.html file which you can find here and then just change it, but you can of course, do any change yeah. So this is just a um just uh something that you are able to see that there has something been changed right.

B

This is a non-breaking.

C

Change this is a non-breaking change exactly because, um as you might know, when you are working with with customers and you're doing breaking changes, and you do not have any pipelines, because um your customer is sitting behind your back with a verbal gun in his hand about fix this now right, then you, you know that um I've.

A

C

So now I'm committing.

D

C

um So, let's look what has changed? I just have changed the index html file, um so let's do a git commit now um bumps version or whatever.

C

Now I do a good push and then, of course, I need to um to use to start the pipeline. This is also something um which I do with um with my script, which is part of that um book example repository.

C

um Why I use script here. I can show you this in a second because easily, um I'm again, I'm using the. um So it's in the context of the service account again, um and um what I'm going to do here is that um yeah that this is executing the build pipeline. uh The bill pipeline in.

D

C

Of that um of that service account, I was creating and unfortunately right now um the ui of openshift doesn't support using or changing the service account, um at least not when I was starting all the demos and stuff like this. So um maybe it has changed now. So now, let's go um to the book ci environment, where my pipelines are so.

B

Yeah, normally it chooses the pipeline service account right by default.

C

Exactly it's the, um but I have um chosen to create my own service account um just because well to show that it's working right and then you have to deal with the um with the other challenges so that I'm not able to say well um just start the pipeline here or um do whatever, but anyway, um in a real environment. I I would say that you would not use start pipeline here, but you also execute it in a different way.

C

So now we have the pipeline run here, which is doing exactly um what I described already. So we are cloning, the source. We are cloning, the conflict. We are packaging, everything um we are building the image now and by the way, um I'm a lazy guy. So, um instead of doing uh of using a docker file behind the scenes here, I'm using what quakers provides us.

C

So I'm using the jib container image builder- and this makes the complete setup of the pipeline so easy, so I'm so. This is amazing task, and this is a main task now. um So this is what I like quite a lot. Oh, it's already done.

A

And you extract from the output, I think this is important if you can also, if you can share them, if you can show us the the the source code, the yaml file of the pipeline, um how do you extract from the output of the quercus build and push right? How do you extract the hash of the container image and giving parameter to the next task?

A

This is also cool to see if you can show us that the tectum part, oh.

C

Yeah, this is the tactile park apart here right now, so what I'm doing is um so extract. The digest is not doing anything else. Then taking um the jib dash image, dot, digits file, which was built during or creating a good time, um and then I am taking this the content of that file and put it into um the tecton task results.

C

A

So basically, you are creating a file with the digits, and then you are sending into result. Result is a uh is the way tecton used to you know to print in, uh to give the an output result that you can use also an input. So this is the powerful thing. No.

B

Yeah, so how do you use it in the next step? uh This would be.

C

The next step, I think, yep, so um here I'm doing um I'm using updated digits, which is um using the customize tool to say edit set image, and um then we are using the current uh wait. Where is it ah here, thomas new digits and the digit is coming ah now? I need to go to the pipeline, of course, the dev pipeline and um the digit is coming just. Let me check here: where is it? Where is it? Where is it um yeah? It's over new ditches? So it's here we go yeah exactly.

C

Can you see it yeah you can um so this is also also automatically being extracted from this, so the parameter um or providing parameters, input output variables from one task to the other in a pipeline. This is quite quite nice.

A

Indeed, indeed- and this is in the source code right- we put in the the caption- uh everyone can also look at the wall source code. In your repository that we put here, yeah.

C

Fantastic, it's everything in there.

B

Right- and I I think, and the image or digest is uh something in the focus properties or something like that. If I, if I'm not.

C

Correct uh yeah, so when you're using jib container build um the quadcos extension when you use this, then um uh when you're pushing the image to a external registry, then the that file is being created so image dot. Digit, I think, is the name.

A

um I know you know jeep and uh there's another tool called j-cube that can do the same thing and also use uh jeep. um I know chip is using kind of uh distro less container image. No, the the container image will be not something like from uh any from fedora from centos. It's gonna be a kind of distroless, uh I'm wondering if you can also control the layers like. If you want to start from a certain, let's say: fedora, can you can you do still do with jeep? Do you know if that is possible? It.

C

Is yeah, so this is um all done within quarkus, so if you want to have a different base image to to build your um your image from then um just have a look at the quakers um container image. So there's a there's, a guide on the quakers.I o website where you are able to see okay. This is the way how I am able to do it.

A

Wonderful, wonderful, this uh so yeah. If you are using qualcomm, this is very lightweight. No, you don't need to use any uh kind of a docker build or or container built in in the node. It can be just quarkus in general. uh The agnostic way would be you know using in our case builder, because it's uh dockerless demo less, but you can use also docker build if you have that multiple way, but this is very very cool.

C

No exactly so, I just wanted to make sure that um that everybody is able to understand that demo right. So this is the reason I'm using jib for this, and um because it's so easy to create your pipeline, then because it's as mentioned just another making goal which you're calling and that's it um so yeah everything needs to uh should already be updated, which means, if I'm now going back to the um deaf book book, dev, environment, okay, well, everything is already updated.

C

So if I now reload, you see ah cool, yes, wow everything is done, so our new application is up and running and, of course, in a staging environment we still have the old one. um So if I go back to staging, if I open this one, then of course we still have um no that's wrong. um That's the book death! This is stage so this one, it's still 1.6 and now we would like to create um to use um the the staging pipeline and again for this. um I again put something based on that.

B

This part is, uh is, can be a bit tricky, so yeah, let's see how we do it, and maybe we can have a conversation on on different uh ways. This can be achieved, especially with tecton, okay,.

C

Well, I'm creating a review, so I'm calling the pipeline script with stage which is and with a parameter, whatever version one dot. What did I say? 0.00.

C

Yeah um and now that is starting the tactile pipeline.

C

um So if I go back to openshift now into the ci environment pipelines, you can see that now the staging pipeline is running and that staging pipeline is doing the following stuff. So, first of all it's cloning, the config, so the person service config repository.

C

Then we are creating a branch um in this person service, config repository um and again we are now extracting the digits. um But this time the digits is coming from the development config now, so I would like to use the current state of the development um environment. I would like to use it um in my um for for staging this yeah. Of course you could also do something like use that digits, for example.

C

In order to do this, but in my opinion, or in my case I was just using well use the the current state and the current development environment to stage this. So um this is one thing. Then we are taking the image using spokeo um on the quail.

C

So if I go to my personal way, I o environment, oops.

A

Yes, scope is a tool to you know: copy container images right instead of doing uh kind of a docker pull or podman pool and podman push docker push. You can use, compare to copy and push right.

C

Exactly exactly so, um scopeyou now has used the um the digits um of the last build to create attack, which is called 1.7.0 tv, and this should also have or would should reflect my um github repository environment here. So you can see that there is a new release branch with that tech, I've, I've created and, of course, right now. Nothing has happened with argo cd because um argo cd is watching in this case on the main branch. um So, of course, in a real environment.

C

You wouldn't use the main branch for this, but the staging branch, or something like this and then the production branch so.

C

So um this is um yeah, but in this case it's the main branch we are using here. So if we go back to um here to the uh staging environment, um we entered to the book stage. You can see here right now. There has nothing changed. So if we are refreshing, this um synchronization is still like the same. So it's still 1.6.0 um life, as you can see here quickly. um So what we need to do now is, of course, in a real environment.

C

You would have a devops engineer or an architect or whatever somebody, um and that somebody is now going to get saying. Well, let's compare what um was changed now we are creating a pull request. Now, um the creating the pull request could also be automated, but- um and there is a gh tool available, which you could use for this um to directly um interact with the github or gitlab um apis, but I haven't used this so far, but maybe in a later release, I will do this yeah.

B

I think we already also provide the cluster tasks for that.

B

C

Awesome yeah- and here you can see now um that there is a change and this is exactly what has changed, and this is again. This is something um what is really important for the git ops paradigma, that you always know what is live and what has changed from the um from the older versions yeah.

C

um So now I can exactly see what has changed the the red one um was the old version and the green one is now the new version um so and if I now, of course, go back to that, one I say well um seems I'm able to merge it. So I'm going to merge my pull request.

C

I confirm the merge and now that it is merged, we are able to go back to the top to sorry, not topology view, but to this one, and it takes one to two minutes round about until argo cd is picking the change up, so I'm shortening this a little bit. So I click on refresh and you see whoa.

C

There was a change now, um please, let's go back to um do a kids clone and um now, let's, let's um publish the changes on the staging environment and if we go back to the staging environment, which is this one um after a while, you see yay.

C

My very important changes last minute bug fix is now up and running yeah. So this is at the end it so there's nothing special in between, but just that um tecton is doing. The build is doing the staging.

D

And then you can also.

C

Do this exactly more or less the same with um a prod pipeline or a pre-pipeline, or something like this um and yeah just use a different different branch, name or whatever? But this is completely up to um how you were defining this and yeah and argo cd is the guy in the middle, which is doing the actual cd part. The deployment part.

A

And and this cd part could be also multi-cluster right, you can have a one argo controlling multiple cluster or one argo per cluster. uh That's that's powerful! In that case, uh if we have multiple clusters, how do you see this structure? Will you have a the central hub arcocity pushing to multiple clusters is still valid. This uh architecture, you show us yeah.

C

So I I can't show it because I just have one intel nfc system, um but um yeah, so this would be exactly the same instead of using just a second, um let's go here instead of using um in cluster. I hope you can see it yeah this url, you are using a different ul yeah, so where your um uh production or testing, openshift, cluster or kubernetes cluster or whatever, is residing and then instead of using this one you're using that other one and rcd is taking care of the deployment.

A

Cool pretty cool um well wow, that was great and it you can see the people can see it's live because I we share in the chat the repository that you were using. So you they. You can see the the pull request you made the the merge. That was all live uh and it's impressive amazing. uh So what is the next step, of course, downloading the book, because you find all those um um all those uh examples there right, uh maybe andrea.

A

We can share again the the link to the book so that people uh can see it uh so yeah. Really our.

D

A

Suggestion is to download the the book um and to find this example and also more content. I've seen in the list there also there was a part about tecton security. No, uh I think, there's more.

C

Yeah, um so I'm talking briefly about tecton security too, so, um especially in the context of um okay. How can I provide my security um credentials to github to create.io, for example, um so this is what I'm going to discuss in there. So really briefly, um if you need more, then um there's a very good, um so the the the documentation.

D

Of tactin is very, very.

C

A

Well, fantastic, uh do you have any final question, andrea jafar, before we close up.

D

One stupid question: actually I do want to disturb the flow we've been seeing. The animation of the pipelines uh on the ui is does tectum uh provide. That was that provided by openshift, this.

C

B

C

D

Upstream texturing will already have a ui.

C

No, no! No! No! No I'm sorry! You are absolutely right. um So um upstream tecton I think, has a dashboard ui, but um that I'm not sure I've never used or um had a look at the dashboard of tecton. But when you're using the openshift pipeline operator, which you can see or sorry.

A

Let's re-share.

C

Your screen yeah, I'm free.

A

Which you can see here.

C

So openshift pipelines um this one provides an overshift developer, console integration, and this is exactly what you see now. So, when I'm going back to my book, ci environment here and having a look at the developer view the pipelines, then you see this one is completely coming from the openshift pipelines operator.

B

Yeah so so I I believe the tecton uh dashboard, like the upstream one, shows you like the tasks and and the logs and results and stuff like that. But it doesn't give you this uh this visual graph and and the the other thing is the editor like if you click on the pipeline itself, and then you go to edit yeah second.

C

B

Actions edit, that's uh one of the like great values you get in here- is that you are able to use the the tasks that you already have in there and visually create your pipeline, which is, I mean very, very useful, because uh pipelines in texan can tend to become very lengthy, and uh you know hundreds of lines of yaml files yeah. So exactly.

C

I would also suggest that when you're, starting with tektin, um that you use the openshift pipelines editor for this with just design your um your pipeline, as you can see it here and then you can always easily go to the yaml view and, for example, just um take that file, and then you store it locally within your project, for example, yeah.

C

So exactly because you can see it here, it is a little bit lengthy and honestly, who said that yaml is a great um configuration technology. I can't understand this okay. Well, um the difference between yaml and xml is also so. Xml is also quite lengthy and um but janu where.

A

C

Are part of the syntax I like it.

A

We're getting into religion battle over here right, yeah,.

A

That's an interesting one but yeah. I think uh it's very cool what we've seen and uh vania. If the people want to try stuff in with reading the book, I think they can download code ready containers, which is a local instance of openshift for free that you can download in your workstation start and openshift locally and start using tech on argo, cd and openshift and do the same example. You do you did it today with the nice pipeline, ui builder and all the stuff right.

C

A

Everything, fantastic fantastic. We put the link in the chat. Please download quality containers to to try it out. If you have any openshift just download the book, read the book and send feedback to vania follow vania on twitter. uh We we have your twitter handle manual. Let me I already shared in the chat. Oh yeah. Here we go, I'm ready to share it, share it in the chat.

B

A

Thank you very much and we we close up folks. uh Thank you, vanna, for this all really awesome. Awesome demo live demo. uh Let's close up today open shift tv. uh We have level up our. We have jafar doing the live demo.

A

Another live demo wow super day and talking about githubs, we have the github's guide to the galaxy on april 21 by our github extraordinary christian hernandez, please in april 21 the birthday of rome also nice day, and please come back here- the openshift tv coffee break, because we will come back next week next wednesday, with a session andrea uh about ci, cd.

D

Yes, uh so next week we're still going to be talking about ci cd, but this time uh we'll be talking about csd pipelines for ai, for ai-based applications that use red hat, open data science and we'll have the one only max murakami.

A

Fantastic fantasy folks, very amazing, show thanks again vania, please download the getting github's book from reddit developer for free and see you next wednesday ciao. Thank you.