Red Hat OpenShift GitOps Guide to the Galaxy | Red Hat Livestreaming, 24 Feb 2022

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: GitOps Guide to the Galaxy (Ep 33) | Exploring Argo Events

Description

Want to know how to build an event-driven workflow with Argo? In this episode we’ll explore using Argo Events with Argo CD Event-based architectures are a key part of building solutions where individual components are decoupled and when we're talking about serverless workloads or Argo workflows, Argo Events is the tool we need for event-driven workflow automation. In case you are not familiar, Argo Events is an event-driven workflow automation framework for Kubernetes. It allows you to trigger lots of different actions such as creating Kubernetes objects, invoking workflows, Serverless workloads, etc., on over 20 event sources such as webhooks, S3 drops, cron schedules, and messaging queues like Kafka, GCP PubPub, SNS, SQS, and more. So join us to find out more and learn more how we can build event-driven workflows with Argo Events.

Follow us:
Christian Hernandez https://twitter.com/christianh814

What is GitOps Guide to the Galaxy?
Every other Thursday at 3pm ET hosts Christian Hernandez and sits down to discuss everything in the GitOps universe, from end-to-end CICD pipelines to creating Git workflows.

Subscribe to Red Hat's YouTube channel: https://www.youtube.com/redhat/?sub_confirmation=1

About OpenShift:
Red Hat OpenShift is an open source container application platform based on the Kubernetes container orchestrator for enterprise application development and deployment.
OpenShift on redhat.com https://red.ht/2o4ccDk
OpenShift on github https://github.com/openshift/

#RedHat #GitOps #Kubernetes

A

A

Let's get into the big topic of open source, something that we actually have an open culture.

A

What is that process that a developer or, let's say, has the kubernetes ecosystem really.

A

All right and we are live- welcome everyone to get off skype to the galaxy I'll, be your captain in this journey, uh christian hernandez, uh technical marketing here at red hat and uh excited I'm excited for uh to today's uh today's episode. We're gonna be talking about argo events right and so trying something new um anytime I put exploring in the title right so exploring argo events. That means that you're going to see me doing a lot of hacking and hopefully breaking some cool things, and also that's going to be your signal to that.

A

We're going to revisit to that topic um part. I was actually talking to stephanie um here, stephanie the producer about the the survey right and uh part of the survey that you guys did by the way. Thank you. We did. We did that last um last year closing last year.

A

I can't believe it's 2022 already, but uh one of the the things that um uh that came out of that survey is you guys said that um you guys would like uh for us to revisit some topics or further expand on some topics that we've already covered right, and so um I decided to kind of do this exploring thing right and so anytime, you see exploring in the title it's going to be kind of your cue of like we're going to learn something new and there's gonna be further episodes as we explore uh further on so um um so the um uh uh today's uh before I get get on to the the topic.

A

A waleed's here welcome ali welcome, welcome here um and uh oh and tiger's on as well tiger the intern, although tiger no longer the intern is here but that'll be forever your name uh for me, um so um so yeah. So before we get into into that topic. I just have a couple things to share with you. um Let me um let me share my screen here, um and so um the first thing I wanted to share with you is um the the vulnerability, the argo c vulnerability.

A

I actually didn't really have much time to talk about it. I think we had a guest last time and like uh before then it like it's been like a month or so right, and so um I actually talked about it on the stack rocks um office hours, which is why I have this up here. So if you haven't seen the stack rocks office hours go ahead and check that out, I did a guest spot there with michael foster, a really cool guy.

A

I told him he's one of my um top three favorite canadians so um he's I I told him that personally so check that out. We talked a little bit about the vulnerability here. I um I kind of want to re um revisit this a little bit maybe go into a bit a little bit more in depth. I will put this um in the chat here, so you guys can uh read it on your own right.

A

um I'm pretty sure you guys have ever read about the argo city vulnerability, but the um the too long didn't read right is the tl, dr, is that um in argo, cd there's a way if you're using helm, charts specifically right.

A

So if you're using helm, charts, please pay attention to this, there's a vulnerability uh that you can actually break out of the application and then and then see other application resources um in um inside of argo, meaning like all right, I can deploy a helm chart and then basically do a cd dot, cd dot city, dot, dot and see everyone else's resource, which is really really bad. So um so here they have. Ironically, they use the red hat logo black.

A

I always I always find that funny, um and so, if you want to read about it, um go ahead, read about it there. The biggest biggest thing is take away from this is please um update your version of argo cd right if you're using openshift get ops, we already have a patch for it, so just make sure that's updated um and that um so that way, you're not vulnerable to the cve.

A

If you're using upstream argo cd just make sure you're using the uh the latest, you know um uh z stream, for whatever version you're using right and so part of the cool thing again with red hat. I I explained this on the the the stack rocks office hours is that we actually um back ported the fix to um the openshift, uh uh get ops version 1.3 right. So, if you're running openshift sorry openshift get ops 1.3 um on openshift version, 4.6 make sure you update that as well. We actually backported that to an older version.

A

um So um so there was oh, so good question rockhound! Thank you for asking. Will there also be a series about stack rocks? Actually, I did have stack rocks. I did have actually michael came on my show um not too long ago. um If you want to catch past episodes, um you can visit this link right here, really cool bitly, that I stole that I still write get out. So I'm like I'm, going to get on this train um that uh red hat's bitly.

A

So if you want to see past episodes, I did do a pass episode um with stack, rocks so um check that out in in the chat um we go well. I can also invite him back on and we're going to go further in depth. um He did a pretty cool demo. He call he called out my applications that had a bunch of vulnerabilities, so I will um which I, which I I I say that I did on purpose so I'll I'll. Stick to that, so um so yeah so anyways.

A

um This vulnerability make sure you're patched make sure um it's pretty cool um that the argus, the argo community in general, not just argo city, but the argo project as as a whole. um You know uh hired these guys to do an assessment right and you know, goes to show that there are. um You know they said, please find vulnerabilities, so we can fix them, which is really cool thing about open source, um and you know shows that they are.

A

They have security first right in the in their in their minds, so um especially important for graduation right for cncf, graduation, um all right cool. So um next up here argo events right, and so I think uh we we talked a lot. We talk a lot about get ops on the show and we touch on um other subjects right that are kind of near near get ups things like cluster management right. We have the acm guys come on. um You know we did a little bit with techton right. We talk about cd.

A

I think um I want to explore more in the argo ecosystem, because there's just so many tools out there that you're actually going to need right, you're not going to only need argo cd right, there's going to be a lot of things right and one of the things is argo events right, and so um one of one of the cool things um that I wanted to talk about is argo events.

A

So here is the link to argo events for the um uh for the official um uh uh documentation for argo events or so so what is what is argo event, so it is. um How do I explain this right, and so um I think I think I want to make an analogy first for for those of you who know me, I I love, I love making analogies right and those of you have been watching the show um a lot.

A

Noah knows, I always say, peanut butter and chocolate right um to make things how two unrelated things can be cool together. So I like making analogies to to make it simple right so before I actually um actually go through the um there is an architecture here that I can't seem to find anymore uh concepts. Is it other architecture, okay, cool? Before I go into this?

A

I I kind of want to make an analogy here is the analogy um that kind of describes, argo, cd events and so um um so think about so I'm gonna go through the kelsey hightower. You know his example of the post office and kubernetes right, and I think I think I'm gonna amend that a little bit and use the post office right and use writing a letter. um It's so archaic right.

A

One day, someone's gonna talk about writing a letter and we're gonna have to explain what that is, um but um but just think about um so y'all. Remember chris short right, so um you know me and chris short, we're we're still friends right. We still keep in touch actually daily. uh We talk daily, um and so, um let's, let's say for a moment right, I'm gonna make the analogy of I'm gonna write chris short a letter right. I want to write him a letter, um so I I write the letter.

A

I um you know I put an envelope, I you know pay the postage and I give it to uh the mail carrier. The mail carrier says all right cool. I got it right and if you, um if you think about it right, um that is uh an asynchronous form of communication right. So all I got was the acknowledgement that the post office has gotten my letter um so whether that gets like lost in the mail, whether that gets. um You know the dog eats it right.

A

Rather whether that's you know in a plane- and you know it got caught under under a seat. I don't know that my letter got lost right. All I got was the acknowledgement that my letter was received, but I don't know that what happens there, like you know, like I said before it could have gotten lost in the mail it could have. uh You know uh chris shorts dog could have eaten it. Chris short could have opened that letter.

A

Chris could open that letter, read it and then throw it away or not even like throw it away, um and so um you know we think about these asynchronous types of communication. um Where I don't, I just get the acknowledgement, but I don't you know we. We always try to build our systems to be synchronous, but like that's not how we, how we work in the world right. That's you know how we we don't. We don't do that in in our world.

A

So um so now imagine the the analogy is like okay, I write chris short a letter and this instructions on how to you know make a cake right. um I don't know how to bake, but let's just follow me on this right. I I can tell him an instruction on how to you know he can make a cake right. So I got an acknowledgement all right. My letter may or may not gotten there. Let's pretend that it did get there and he reads it and he makes the cake right.

A

So my my event right triggered him to make a cake. I didn't know whether not whether or not he did it unless he called me and said. Thank you, um but you know those two actions are decoupled right and those two, um those two actions are um um are are not um not related to each other they're asynchronous and they're, like really really like different from each other right and so um rockhound. I see her.

A

I see your question um I'll answer that in a little bit right, and so let me let me kind of just finish this, this analogy, and so, um and so the idea is that this is kind of what argo events is the best way. I can explain it. My analogy didn't make sense, go ahead and reach me out on twitter. I will um I'll try to find a better analogy, but it's essentially, I write a letter how to make a cake. Chris short gets the letter reads and it makes the cake.

A

I don't really know that whether or not he did that and so um and um the and that's kind of like how argo events is right. So rockhound asked when is arguably events we plan to be part of openshift get off's operator.

A

um It is, we are still evaluating um if we're going to include argo events right and so um there's always an evaluation phase. um You know with with with things you know, being red hat and being in part of open source like it's hard to include everything right, so we're still like in the evaluating phase right um in terms of if we're gonna include argo or argo events, um and so that that's where we are sorry, I don't have a better answer, um but it is.

A

It is like something that's on our radar, so um so, let's go over the architecture, really quick right, I'm going to open this image in a new tab, because I want to make this big, so you guys can see right. I'm going to flip flip flipping being I'm going to be flipping back and forth between this. uh You know these two tabs so I'll. Try not to do it too, quick. So um s do not.

A

um um uh You know, make you guys dizzy or anything right, um and so here so there's there's there's a um there's. I guess three main parts of of argos, cd events or stars. Argo events and our go cd events so um out here is um an event right and so what it most people when they think about an event they think about like a web hook right- and I think- and that is an event right, but it doesn't necessarily have to be an event it can be, um it could be a web hook.

A

It could be um a uh you know, an s3 bucket right like something something happened in s3 bucket, a gcp pub sub. It could be an amq message, it can be, it could be various things right and if it's just some sort of event something happened, it could be something as simple as someone wrote a file somewhere right like someone touched a file somewhere, that's an event right and so um um argo city.

A

Has this thing: it's like a it's like an event listener right, and so it listens for all these events, and so um you know it, it basically is listening for something to happen um and when something does happen, it'll um it'll write it to what they call the the event bus right. So the event bus, all all the event bus is, is a wave.

A

It's a communication, it's a highway right, so it says basically an event happening and the in in argo events will say: okay I'll write this to the event bus and it writes it to the event bus um and then the next part is that there is a sensor right and so sensors um sensors kind of made up of multiple parts is like a like a sensor and then there's a trigger that comes after this, but I'll explain. First, what a sensor is right.

A

So a sensor, um all it's doing is is a reading from the bus events that happen right um and then from there. It can trigger things right so first, for instance, it can trigger um an argo cd um uh rollout right of like you know, or it can do a workflow or it can do um uh you know it can interact with your kubernetes cluster. It can do something right. A trigger is like doing something right. What's really cool is that the event what's happening is decoupled from what's being triggered right.

A

So that's and- and the reason for that is the reason kind of what I'm really excited about. Argo events is that um in the sensor there are dependencies that you can do right. So, for example, not just um not just reading, not so like. Normally it's like a like an action reaction, sort of thing right, like I hit a hook, something happens but with since they're decoupled, you can say I'm waiting for three events before I trigger anything.

A

So, for example, this is a dumb example, but I'll just do it anyways, because it's a simple one like I do a um you know, a git commit right to a specific branch. So that's an event and someone does on on slack. You know, uh chat ops. They do a slash test right, so those two events need to happen in order for me to trigger.

A

um uh I know um I guess in order for the the sensor to trigger whatever trigger that that I set up so um so so yeah, so that's kind of like the high level. I'm gonna be going again going back and forth because we're going to be looking at um different kinds of event, event, sources um uh and different kinds of triggers and sensors as well.

A

So um so that's kind of like the high level. I'm actually I I learn by doing right, so I'm actually going to be doing this thing here um and so everything that I that I'm going to be doing is, in my repo, so um um I'll I'll paste that in a little bit- and um you can do it- um um you know at your own time or if you want to follow along- maybe maybe I'll post it a little bit, but so first things.

A

First, I need a kubernetes cluster right so for for those of you, um I'm not going to be doing this on openshift yet yet because I didn't really have a a good use case for it. I will maybe next time we we we we play with our go events, um so I'm going to do any kubernetes cluster, so I'm going to do um oops um if you haven't used kind.

A

It's like the coolest thing ever you can get a kubernetes cluster in um uh use this image there we go um like in a few seconds right, and so um actually he did have surgery today, yeah he he talked to me um um today, yeah so um yeah, it's almost like uh yeah, um while they'd. Actually so we'll we'll um we'll talk about the serverless component of it as well right, and so we have um so now we have a cluster right and so let's do k get nodes cool.

A

um So we got it's not ready, come on.

A

Yes, let's break everything.

A

As I was testing, this watch docker rate limit me right as I uh okay there we go. I said, watch watch docker rate limit me because, um because I was testing all day today, uh let's get nodes very cool, so um first things. First, uh we need to create a namespace.

A

I'm gonna create argo cd events right. I created a namespace to house all of this thing here. So then I'll install um install the events yaml and since I can't can't memorize urls anyone could memorize urls my hat's off to you uh first to create that right, the the installation and then the um components here like crds and whatnot uh cool. So all right. So let's take a look at.

A

Right and so then, basically you got the the event bus right, like I mentioned over there in, in the other graphic right, the event, bus, um the event source controller right, that's right here and the sensor controller. That's right here right, so you got those three controllers, all doing their their thing um cool. So then, let's let's, hopefully that created all right. If docker right limits me, um hopefully they don't docker hub, it's all prey to the docker docker help of gods.

A

Here, uh let's look at the event source right and so um uh the event source here is uh a web hook by the way, um oh, I'm gonna butcher his name. I am completely sorry. I wrote it down, okay, so victor. So if you guys haven't seen victor's far six, um uh um he does this like whole series on argo events check that out. Definitely great thing shout out to him for um um for putting on such a great video series, so check him out.

A

um Victor um farsik he's uh developer advocate for upbound, so maybe I'll have the upbound guys one day here um on the show so back to topic at hand so event source right. So this here, um I'm basically I I said that this could be anything um not just a web hook, but here I am doing a web hook, but the web hook is kind of like the easiest thing and the easiest thing just for someone to understand.

A

What's going on right, so here I'm going to create a web hook and that's going to be my my event source. So if, um if I go, um let me go back to my web yeah here. uh So event starts so again. Event source could be um all kinds of things right, so it can be like um aws sns amq um calendar, which means like a like a cron. Almost um it could be kafka. It could be min io right like when something happens in a bucket. um It could be a slack right.

A

Remember we're saying, chat ops, so you do slash something else right. um You know, there's netapp storage grid, all kinds of stuff right and so, but we're doing web hooks. um But these are just like events. That's happening outside um outside your cluster, and so what we're gonna do is uh uh create a web hook um or call hello world right, we'll and we'll just. We will just listen for post for hello world right and so let's create this here. Okay, apply.

A

Demo, that's right event, source all right cool created, so let's do a get event uh come on contact event, uh sources, um argo events, and so uh so here so it created the um the the event source web web hook here and then let me check my notes, yeah and then so. um Okay get a pods and service on argo events right and so, as you see here, you actually see this web hook uh resource.

A

um So it actually created a pod to listen for uh that web hook and so and then it created um uh the service right to listen on and so for a really quick demo, I'm going to open up a new tab and do a cube, ctl, argo, cd event. Sorry, argo events, um I'm so used to saying argo cd. I don't know about um once you're like embedded in the world. It's almost like muscle memory, um uh oh yeah port forward and then I'm gonna do svc.

A

uh What's the name of the service.

A

I just copy pasta that, right by type when you can copy paste um so event source and then was it: uh oh yeah, two thousand twelve thousand right cool. So um the reason we poor forwarding is because we want want to be able to hit that web hook, and so um so we have this. uh So now it's it's listening for things, um and so let's do uh let's clear this, so you guys can see it. Let's curl it right! Oops pearl!

A

uh I always have to look this up. Post uh h, content uh type, application, json right so like you're, just gonna do a regular standard web hook um and then this here uh my payload is gonna. Be I'm gonna do a message this could be like you know it.

A

There's um I'm just sending it a json payload I get. This could be formatted. However, you want right, but I'm just gonna say I'm gonna send a message: I'm gonna do hey there um and then you do json on the fly we see we'll see and then we're gonna do that http 127.0.

A

Maybe someone can explain this to me when I do localhost this doesn't work. I have to actually put one two. Seven, zero zero online, so if anyone's smarter than me um can explain why we'll see so um yeah, so we did uh hello world. So if you uh I shouldn't, have cleared the screen but I'll we'll look at it after, um if you remember that was the end point that I put in in the web hook right and so notice here um it says success right, handling connection, but that's all I get right like that.

A

It's non-blocking right it received my my message: it wrote it to the bus and that's it that's all the event sensor does so um uh even source does right so here. So here I I did. You know 12 000 slash hello world, which is because that's what I'm listening for and I did a post um and then the payload I just sent it some json. So um so that's um and then it just it just says all right cool. I I got it.

A

I got it that's kind of like the the mailman, sorry, the mail carrier. um Taking my letter and saying you got it. That was it right. I don't know what happened to my um uh my uh my event. There I just created an event. So um so what happened? There is that it took it and I wrote it to the bus and that's it that's it's um it's job.

A

It's done is for for the event, source, right and so um what's really cool now is now we're going to create a sensor and that sensor is going to trigger something. So let's take a look here um uh actually, I think I have to vi it here uh so sensor yeah, because it's kind of it's a big guy. So here um I want you to pay attention to the lines these lines here, eight through eleven. um So you know I create a center right. um I named it web hook.

A

It doesn't have to be named the exact same thing. It could be named whatever you want, um but here dependencies it says. Okay, so um the name right name something. You know this is a payload I got it can be whatever you want, but the this is where it matters right. You need to have a web hook, that's the source name! Remember we we uh the event source name, we called it webhook and the event name. We called it hello world.

A

um So if you, if we get out of this- and um so you know the name web hook and the web hook, name is hello world, so that's how those um those two things match.

A

So we have those so notice that this here is an array right, and so you can have um multiple dependencies, meaning you can wait for you know um not just one event but like a group of events right so like. Maybe you need a few things to happen before you trigger something so um and when I say trigger something. That's literally here. What I mean here is, I have a trigger or they call a trigger template.

A

Is that any time the dependencies are satisfied?

A

I want you to trigger this, and so this here is essentially I'm just going to create a pod right, just kind of something simple um right here I say uh it's a trigger template, I'm going to create kate's right. This is the the gbk right group version kind.

A

I want you to create it right. You can do create delete right patch whatever, um and this is here. If you recognize this, this is like the actual, odd definition, um and so what other triggers are there?

A

Let's take a look, so that's, um let's visit sensor, real quick, just to kind of go over what I said, but sensors defines a set of event dependencies so remember when I said that you can create dependencies between events and triggers and a trick um and multiple triggers by the way, not just not just one trigger so, but what kind of triggers are there right?

A

You can do a lambda function right, you can do you can trigger an argo rollout which is pretty cool right, and so, if someone hits a web web hook, you can trigger an argo um roll out. You can do an argo workflow right, so you can do a workflow like if you're doing some sort of um like pipelining, um you know build your own right, bring your own, bring your own beer sort of situation here.

A

um You know kafka message: uh gnats right, you can do a serverless function, um so wally, that's the kind of like the idea. Is you can do a um you know, create kubernetes objects right so, like there's different trigger types that uh you can do so you can do these things here.

A

So this is kind of um what that trigger is right, like it does something. So so we wrote an event event. Bus um takes that event. Sorry, the the the event source wrote, wrote that event the event bus, you know, writes it to this bus. Now. The sensor here that we're going to create is going to sense for that right. So anytime, someone hits that um um that end. Point that end point we're going to create a pod right. What's really cool about creating this pod?

A

Is this um section right here parameters of the trigger template, but here you notice here it says um you know um we're going to get data right, so we're going to have data from the payload right so from from the payload right, we're going to grab it from the message right so like. If, for those of you who are familiar with the web, hooks right, it's body messages that message equals thing that we have there and um and then here the destination is spec containers. Zero args one so basically spec containers right.

A

This is the first one, so it's zero and then arguments and then one right. So this is zero, and this is one so what's going to happen, is that when I hit that web hook with the message it's going to take that payload and inject it into I'm pointing at the screen like you guys, can see it's gonna inject it into this. This area right here, and so this is.

A

um This is another way where argo uh uh events is really cool, not just for the dependencies, but that you can pass data back and forth and you can pass data back and forth and it's decoupled right, so um we're going back to the uh to the event source rockcount. I see your question I'll get to it in a second. um um uh The event source here is um this here, like let's say like like github it doesn't it get. The github web hook doesn't have to know necessarily what happens at the at the end.

A

Point right, like those are decoupled right, so the the source, this so the event and and the what it triggers is: they're, they're, independent from each other right so they're decoupled, meaning that um you don't have to write a lot of custom code in order to um make your applications talk to each other almost right. So this is kind of um um really really cool. So um question here is a sensor, a global thing and I define limit the triggered event for the pod to be started in a specific okay yeah.

A

So um so this pod yeah since I didn't- uh and I don't have like a metadata, namespace um and actually uh well- it's not going to be created in the default name space. What it's going to be created is in the um the namespace I defined when they, uh when I create the sensor. So wherever this sensor is created is where the pod is going to be created. um So you'll you'll see that you'll see that in a second, so um yeah.

A

So let's, um let's create this right and so kind of just to recap: I'm going to write something here. I'm going to write the uh I'm going to hit the web hook. The event source controller is going to write that payload into the bus.

A

The sensor is going to say: oh hey, someone satisfied this dependency um and then so, since someone says satisfied dependency, I'm going to take that payload and inject that pod, with whatever message I get and create a pod right, and so um let's create that here uh says key kubernetes apply uh argo events, demo, oops, oops, f, demo, uh sensor, right and so um going back to your question is that since I created the sensor in argo events, that's where the pods gonna end up um cool. So then I have.

A

uh I have the port forwarding still going. So let's do uh the curl one more time. So this here is uh I'm sending a message saying: hey there. Let's do um another message: let's do uh you know hello, rockhound!

A

Thank you for joining the stream you're part of the show. Now um yeah we lost the first message correct, um since there was nothing to pick it up. So it's not like an amq stream or something like that. Where there's a retries there might be retries. So that's that's! uh That's a good thing to investigate, um so, let's do it says success right. I received the message here and if I do an oc get pod oops say I'm already an oc um again muscle memory I mean it would work. Watch you'll, see you get pods.

A

um You know, it'll still work right. It's just still. um The payload is right there right. It's completed right. So let's do a okay logs of the payload and then argo events there. It is hello, rockhound, right and so um the uh so. The again you see that what would end up happening is the event happened. The event source controller wrote that to the bus and um the sensor said: oh hey, you satisfied a dependency.

A

Let me take that payload and write it into the pod right, and so you can see how you can start chaining different things together right, so that you can take um like a cut like a get a push to a specific branch and then pass that data to like a workflow like argo, workflow right and start your ci cd pipeline or whatever else right.

A

Whatever other event that you, that you want to see right so kind of going back to the analogy of the um um post office right so like the message got lost in the mail right. So like like the first message I sent the letter, the cake was never made right, and so um you know I sent another letter.

A

It got to it and then chris short created the made the cake baked the cake for me. So um um because you know, and and whatever happened in between that was just uh as soon as I sent that letter and and uh the postal carrier said that he got it and so the um uh um that's um and then the the payload created right, and I can run this multiple times right. So then, if I curl it again- um and I can say you know, make a cake right again. If I do.

A

I do get pods right, uh which one, oh, this one's, the newest one here. um Actually, let's try this uh create. Is it metadata metadata creation times? Damn I'm doing this from memory, ah you can do it that way, cool. So um if it puts it at the end interesting, I would think that the newer one goes on top.

A

Okay, logs, you argo events and then you know, payload receive make a cake right, and so um you can see how um these are um uh how these are um decoupled right and so like. If I do an oc oops, if I do okay get pods, I go events you can see. If I do. Okay logs.

A

Follow uh where's?

A

uh Is it the event? Let's look at. Let's look at all of these guys right. So, let's look at the event. Bus controller.

A

Nothing there, that is to just start basic startup stuff.

A

A

The event source controller uh web hook event, so it looks like yeah, so it looks like the um it logs when the event comes um and then the sensor controller. This is the one I'm curious about. So let's take a look at that.

A

uh Oh, come on.

A

Yeah, so this is that guy there, let's look at the pods again.

A

uh Webhook sensor yeah, so the actual sensor.

A

Now one of these tells you um when it when it got hit here right, trigger print payout trigger name payload trigger. So you can see the those coming through. So rockhound is um got a question here. Can I specify at least time on the bus that the events need to be in the time frame or something else? Maybe asynchronous issues may be handling, or can I just see the bus content and sources for logging track and who? What triggered the events? That's actually a good question.

A

Let's look to see if um uh so like just go to the event bus and see if we can find that. um So it's a nat streaming.

A

That uh events namespace okay, let's see if we can so I guess you can uh modify the event. Bus looks like you, can store them right um or else like they're they're gone right, and so um no, it's like the persistence, whether or not to use persistent volume max age. So it looks like there is a max age and uh max messages right, and so you can do like um um max messages here um max max payload right number of bytes right, so you can control some of those there.

A

um You know priority right, oh interesting, oh probably the staple set of the pods never mind, um and so so yeah it looks like you.

A

Can you have some control on whether or not you can um you know the like the max age number of messages, how big the bytes that would master number of subscriptions right um and so um and as far as as far as tracking, I would think logging would be some of that or I'm pretty sure you can connect um like something like prometheus to it or grafana, to be able to to see what what's going on here um here uh this documentation.

A

So you know now, I'm just basically just going through the documentation of how um how this is you guys could check that out yourself here, but you can have dependencies and you can have multiple triggers. So that's what I really think is cool about. Argo events is that you can have um uh you know they're decoupled right, so you can have multiple events and multiple triggers and you can kind of mix and match how things how things go.

A

um I can see this being very powerful, especially like in ci cd situations, where you kind of have a way to um you know map um like a get commit to let's say like a tecton pipeline um or a workflow right, whichever right here, red hat, we're all in on tecton. You do kind of like a like a like a tecton pipeline um on you know, or multiple pipelines. Right like you can have that you know, since you can do multiple triggers, you can have.

A

You know fire off multiple pipelines uh to be able to do some of this stuff here, um and so I think, the the next time I I do something with with um with argo events. I'm gonna do something like that. I think that'll be really cool to be able to see how you can use some of these things together.

A

Right, um argo, cd, uh you can do um not argo cd, argo, workflows, right image, updater um and um that that sort of uh that that sort of thing right- and so um this is actually pretty cool, um there's a high availability. I like how this is like in bolt and bold is like the first thing I saw. I don't know if you guys saw.

A

The first thing is like the first thing I saw do not manually scale out the replicas that might cause unexpected behaviors, and so um uh so I in anyways, so you guys can uh check out some of these things here. Hopefully this was kind of like a helpful um like a helpful at least introduction.

A

um You know when I first looked at argo events. I had no idea what what was going on, um and so someone once once asked me like.

A

There should be like a generic um way to manage web hooks, and I think, like this is like really close to that right, like a generic kind of like a entry point to manage like web hooks and be able to, um you know, connect things with, while, while still being loosely coupled right, like there's no hard dependencies on on interaction between those right like like, uh um like trying to write an integration between s3 and slack by yourself would be really hard, whereas you can use just argo events to kind of um use that, as like a like a traffic manager for you.

A

So um let's see here, let's see the rock hound says feel like a bit like, like salt stack events and reactor stuff, I actually haven't, played with salt stack right. So back back in my day right, I feel like an old man already back in you know, back we didn't even have ansible right. I was using a lot of chef. um I know salsa has been around for a while, but um I use chef right.

A

So I was, I was a big chef guy um played with puppet a little bit, but um I I I was always I guess I always have to be different right now. This I played with chef, I think like chef was.

A

um It was one of those things here um yeah, so while they definitely definitely right right like so like I've gone through he so while he said it would be interesting to see real use cases and how it's used um so, like um you know this um playing around with this right, like kind of showing you guys how I played around with this, um it was really cool introduction right.

A

It was really awesome, but um uh but I I want to see some of these things connected right like I want to see all of these things working together to get like this. This workflow, I think, um let me make this a little bigger the um you know seeing some of this stuff right, seeing like how a git you know s3 bucket, you know all these things, how you can integrate all these things together, um so um chef was puppet plus plus, yes, which, which is why I liked it.

A

Why I like chef, I started with puppet and then, like chef came out. I was like oh, this is awesome I actually over the weekend, so this is way off topic, but over the weekend- and this shows like I came from a small company before red hat over the weekend- I switched everything from puppet to chef and we just started. You know it was literally like a two-day migration. It's like we switched everything over. Does everything work, yep, all right cool and we started using chef from then on out, so um so uh cool cool.

A

So um again, this was kind of like an exploratory stream right um kind of just kind of getting our feet with argo. Argo events um I want to uh you know for further episodes and other episodes uh we'll do more and more with it. um Other episodes I'm actually going to start. Looking more into the argo project, ecosystem um things like the the image updater I know there was.

A

I get questions about the image, updater, argo image, updater, um argo notifications um and- um and you know, with red hat- this actually made me think of something else. But our notifications is another thing: uh argo notifications, but if you haven't heard argo cd version, 2.3 release candidate is out and when 2.3 comes out, application sets and argo um notifications is going to be part of the core argo cd right, so you no longer have to install it separately. So we'll do a whole stream about that.

A

How how that all that um works so um see here so um see here unless there's any other questions here. uh Thank you ali. Thank you appreciate it um here. Let's see here, there's no more questions so cool um all right, so um yeah. So please, remember! Yeah like subscribe. um Re-Share follow us on twitter. Follow me on twitter. um We appreciate it right. We we appreciate the love here. I I really appreciate it um so um rock hound.

A

I will be in valencia for kubecon, get opsicon right I'll, be a get ops con for sure. um So, um we'll you know I'll see, hopefully see some of you out there right. We get to um hang out in spain um and yeah. So um I just want to thank you. Everyone thank you for for watching, um get ops guide to the galaxy um next time. We're gonna get um someone from codefresh right. We're gonna have hannah one of the developer.

A

um um The developer advocates from code fresh hannah to come to talk about um argo um autopilot right. So autopilot is uh something that's really really cool we're. um You know we actually been working together uh me and hannah over at code fresh to try to get something really cool for you, guys uh next time so be sure to tune in, for that so um other than that um hope, everyone's uh out there being safe and, um like I said always say, um unless it's in git, it's only rumored. Thank you. Everyone cheers.

A