WEBVTT 00:00.000 --> 00:09.000 So we're going to have Mario talking to us, but we're playing with Q, being 00:09.000 --> 00:10.000 pardoned. 00:10.000 --> 00:11.000 Thank you. 00:11.000 --> 00:12.000 Thank you very much. 00:12.000 --> 00:13.000 Thank you, everyone. 00:13.000 --> 00:15.000 So my name is Mario. 00:15.000 --> 00:20.000 I recently joined the Department team at Reduct. 00:20.000 --> 00:22.000 So a few months ago. 00:22.000 --> 00:26.000 And today I'm going to talk about something that I didn't know before 00:26.000 --> 00:33.000 that I think it's interesting, especially if you use Q-branadis. 00:33.000 --> 00:41.000 You may not like to have to set up clusters often. 00:41.000 --> 00:43.000 So let's start. 00:43.000 --> 00:44.000 This is the agenda. 00:44.000 --> 00:48.000 Yeah, we'll briefly talk about pardon and Q-branadis pods. 00:48.000 --> 00:50.000 And then pardon and Q-play. 00:50.000 --> 00:55.000 That is the pardon and command that is actually the main topic of the talk. 00:55.000 --> 01:03.000 So first of all, let's briefly talk about pardon. 01:03.000 --> 01:06.000 So we are as fast and container room. 01:06.000 --> 01:11.000 I don't think we need to talk too much to an introduction about pardon. 01:11.000 --> 01:13.000 So I'm not going to do that. 01:13.000 --> 01:18.000 I assume that most of the people know what pardon is. 01:18.000 --> 01:22.000 I'm just going to provide some information. 01:22.000 --> 01:28.000 So it has been released in 2018 for the first time. 01:28.000 --> 01:35.000 And then it's mostly maintained by engineers from Reduct. 01:35.000 --> 01:41.000 We have a lot of contributions, but the maintainers are mainly redactors. 01:41.000 --> 01:49.000 But recently, so a few months ago, we have proposed to have pardon 01:49.000 --> 01:51.000 as a CNCF project. 01:51.000 --> 01:55.000 And a few weeks ago, it has been accepted. 01:55.000 --> 02:01.000 So this is also because we think that in a vendor, 02:01.000 --> 02:05.000 a neutral organization, pardon, may grow more. 02:05.000 --> 02:08.000 We may have more companies contributing. 02:08.000 --> 02:17.000 We don't want for that to be the only that is governing the project. 02:17.000 --> 02:25.000 So now a few comments of pardon, just to try my slides are actually working. 02:25.000 --> 02:28.000 And also to talk about pardon briefly. 02:28.000 --> 02:30.000 So just a few comments. 02:30.000 --> 02:35.000 The first comment that I run to reset to make sure that the next comment will actually work. 02:35.000 --> 02:41.000 So I'm cleaning up all my containers. 02:41.000 --> 02:46.000 Let me make it a little leak. 02:46.000 --> 02:49.000 Yeah. 02:49.000 --> 02:58.000 Then I will just show a container, just the SHVD image. 02:58.000 --> 03:04.000 And to make sure that it's working, I'm actually going to run curl. 03:04.000 --> 03:07.000 And yeah, it looks like it provides, it works. 03:07.000 --> 03:11.000 So these systems slides, so I will use it. 03:11.000 --> 03:18.000 And these are actually comments that are executed on my machine. 03:18.000 --> 03:25.000 That's a Linux machine. 03:25.000 --> 03:29.000 So let's talk about Kubernetes pod now. 03:29.000 --> 03:40.000 Reality is that even if we like to do podman run into actually manage containers with podman or other containers engines, 03:40.000 --> 03:50.000 we usually have to deal with those emails and to run to test our applications using pods, 03:50.000 --> 03:52.000 the Phoenician or the deployment, et cetera. 03:52.000 --> 03:55.000 And the problem of that is that you need to find a cluster. 03:55.000 --> 04:03.000 So either you run mini-cube or kind locally or you need to actually provision a cluster. 04:04.000 --> 04:12.000 Those procedures may be a don't thing, especially if somebody hasn't been doing that a lot. 04:12.000 --> 04:22.000 Also having a mini-cube close to podman or to Docker on your laptop may start things. 04:22.000 --> 04:29.000 You start having a lot of virtual machines if you are running that on my Quest or on Windows. 04:29.000 --> 04:35.000 The idea is we would like to take those definitions of those objects. 04:35.000 --> 04:38.000 You run at these objects. In this case it's a pod. 04:38.000 --> 04:43.000 And there are a bunch of containers defined in a year. 04:43.000 --> 04:46.000 Let me show it fully. 04:46.000 --> 04:49.000 So first container and it is a ready container. 04:49.000 --> 04:53.000 A second container is the web application and then there is also a volume. 04:53.000 --> 04:59.000 So this is something that usually I test with a cube CTL applied. 04:59.000 --> 05:12.000 So the idea is that with podman we should be able to actually take the definition of those objects and run them locally on the machine. 05:12.000 --> 05:19.000 So in a sort podman is translating that KML in local containers. 05:19.000 --> 05:22.000 So that's the common podman you play. 05:22.000 --> 05:26.000 So let's do a demo. 05:26.000 --> 05:28.000 So first of all let me clean. 05:28.000 --> 05:30.000 I'm cleaning containers. 05:30.000 --> 05:32.000 I'm cleaning also volumes. 05:32.000 --> 05:36.000 And now I'm going to run a Q play. 05:36.000 --> 05:42.000 The pod ML is exactly the same pod ML that I showed before. 05:42.000 --> 05:50.000 Okay. And if we make it. 05:50.000 --> 05:54.000 Yeah. 05:54.000 --> 05:59.000 There is actually a podman out which shows that there is a pod that has been created. 05:59.000 --> 06:01.000 And then two containers. 06:01.000 --> 06:04.000 The containers are the containers we are familiar with. 06:04.000 --> 06:08.000 So the two containers that were the final year. 06:08.000 --> 06:12.000 And the pod is actually an object that podman. 06:12.000 --> 06:17.000 So it's another container that actually we call it infrastructure container. 06:17.000 --> 06:23.000 That's set up the namespaces and C groups for the other containers that will actually be executed. 06:23.000 --> 06:28.000 So the little difference is that to support the pod object. 06:28.000 --> 06:33.000 We need to start first an infrastructure container that does nothing. 06:33.000 --> 06:36.000 Meaning that it doesn't run workloads. 06:36.000 --> 06:41.000 But it's actually setting up namespaces and C groups so that those can be reduced. 06:41.000 --> 06:45.000 By the containers that will be added to the pod. 06:45.000 --> 06:49.000 So if we do now a podman, yes. 06:49.000 --> 06:52.000 Let's see what we have. 06:52.000 --> 06:54.000 So we have a first. 06:54.000 --> 06:58.000 I'm just showing the images here of the three containers that are running. 06:58.000 --> 07:02.000 And I see that the first one is a podman post something. 07:02.000 --> 07:05.000 This is the infrastructure container I was looking about. 07:05.000 --> 07:08.000 And then there are the two ready and the web application. 07:08.000 --> 07:10.000 The two container that I was expecting. 07:10.000 --> 07:15.000 So actually podman has been able to start that pod. 07:15.000 --> 07:20.000 Yeah, I'm locally and I'm going to be able to test it right now. 07:20.000 --> 07:25.000 Another thing is that I expect the volume to be created. 07:25.000 --> 07:27.000 And yes, create the volume. 07:27.000 --> 07:34.000 So that's the state of a ready in this case of the container that is using the volume. 07:34.000 --> 07:41.000 In the pod, yeah, it's actually persisted locally. 07:41.000 --> 07:47.000 And if, finally, I try to test the application. 07:47.000 --> 07:54.000 In this case, it's just the application that does a counter of how many visits it has done. 07:54.000 --> 07:55.000 We have done. 07:55.000 --> 08:02.000 And yeah, you see that it's actually, yeah, if I make it. 08:03.000 --> 08:11.000 Even smaller, I should be able to execute that. 08:11.000 --> 08:16.000 You see that the counter continues, et cetera, et cetera. 08:16.000 --> 08:18.000 And this is cool. 08:18.000 --> 08:25.000 But it can also actually, since we're running locally. 08:25.000 --> 08:29.000 And since podman is able to build images from Docker files, 08:29.000 --> 08:35.000 we may want to actually, and we may are in development. 08:35.000 --> 08:40.000 We may want to build the images before we actually start containers. 08:40.000 --> 08:45.000 And this is an option, a build option that the podman, 08:45.000 --> 08:50.000 Q play, the command, Q play has. 08:50.000 --> 08:56.000 So let's clean everything up before doing the demo. 08:56.000 --> 09:02.000 And what I'm doing now is that I am removing from my local machine, 09:02.000 --> 09:06.000 the image that is used on in the pod, the ML. 09:06.000 --> 09:11.000 So that I'm sure that we need to rebuild it otherwise pod, the ML will, 09:11.000 --> 09:17.000 so podman won't be able to start the container. 09:17.000 --> 09:19.000 Okay, so that has been done. 09:19.000 --> 09:22.000 And now I'm running, I'm going to run pod, Q play. 09:22.000 --> 09:30.000 We'll have the same pod, the ML is before, but we have the build option. 09:30.000 --> 09:36.000 Okay, you see that the first thing that it does, it starts actually doing a build. 09:36.000 --> 09:43.000 And you need to make it really small to actually see all the output. 09:43.000 --> 09:53.000 But I should have completed successfully, and I can check now. 09:53.000 --> 10:04.000 Now if I do, if I run this podman images, I should see that the image that I just removed now is back, 10:04.000 --> 10:08.000 because I just really, let's test it. 10:09.000 --> 10:13.000 Okay, so it any says, it has been created three seconds ago. 10:13.000 --> 10:24.000 So the command podman Q play as build the Docker file before, to be able to do that, 10:24.000 --> 10:29.000 there should be a folder with the name of the image where the Docker file is. 10:29.000 --> 10:34.000 That's how podman works. 10:34.000 --> 10:40.000 That's what podman expects to be able to build an image. 10:40.000 --> 10:54.000 So you provide the build parameter, then it checks if a folder with the name of the image exist with a Docker file. 10:54.000 --> 10:58.000 And if it funds that, it tries to build that. 10:58.000 --> 11:04.000 Once you have a build that is started the container, it's worth the new image that it has just built. 11:04.000 --> 11:14.000 So if we run PS, we see again the same container, so our pod has been started successfully locally. 11:14.000 --> 11:23.000 And we should be also able to see the volume and. 11:23.000 --> 11:26.000 Yeah, and the counter again. 11:26.000 --> 11:31.000 So the application, I'm able to curl, so I'm able to actually access the application. 11:31.000 --> 11:36.000 All right, so this is was podman Q play. 11:36.000 --> 11:42.000 Now, we rapidly, another comment that is Q generate, that is actually the opposite. 11:42.000 --> 11:48.000 So what we do is that we are going to clean the system again, remove all the containers. 11:48.000 --> 11:56.000 And I'm going to do to run a simple simple HTTP server. 11:56.000 --> 11:59.000 It started, okay. 11:59.000 --> 12:04.000 And now, yeah, if I do PS, I see that it actually, it's the HTTP. 12:04.000 --> 12:07.000 The container is running. 12:07.000 --> 12:13.000 And I can run this comment, generate. 12:13.000 --> 12:15.000 Let me be bigger. 12:15.000 --> 12:23.000 So it takes as a parameter the idea of the of the container that I just started. 12:23.000 --> 12:29.000 And if I execute it, it will output the pod definition of the container. 12:29.000 --> 12:36.000 So I can pipe the output and I can to to keep CTL and I can just deploy to Kubernetes. 12:36.000 --> 12:38.000 So these are another feature. 12:38.000 --> 12:44.000 So as though they're way around, so from local container to pods in Kubernetes. 12:44.000 --> 12:52.000 All right, so these are the different objects that are supported right now. 12:52.000 --> 12:56.000 So we support pods, we have seen that, we support deployments, 12:56.000 --> 13:00.000 persistent volume claims, config maps, secrets, 13:00.000 --> 13:03.000 demo set and jobs. 13:03.000 --> 13:07.000 So I showed something that I did a new before joining the team. 13:07.000 --> 13:13.000 There are other things that are really awesome of podmen that not everybody knows. 13:13.000 --> 13:15.000 Routeless a demo less. 13:15.000 --> 13:16.000 This is really popular. 13:16.000 --> 13:22.000 I think most of the people know that podmen can be demo less and can be roofless. 13:22.000 --> 13:27.000 But maybe everybody knows about build farms, image volumes. 13:27.000 --> 13:31.000 So you can use images, mount images, input as volumes. 13:31.000 --> 13:35.000 Podlets, podmen is age. 13:35.000 --> 13:40.000 Those are like, yeah, the interesting things that you may not know. 13:40.000 --> 13:46.000 And finally, just a few resources for if you want to learn more about it. 13:46.000 --> 13:53.000 And there is a talk about podlets in a few hours here from Axel. 13:53.000 --> 14:00.000 So don't miss it because it will talk about another cool feature of podmen that is not super well known. 14:00.000 --> 14:04.000 And I think we may have two minutes for questions. 14:15.000 --> 14:20.000 You can probably do one question if someone has something. 14:21.000 --> 14:23.000 You can take those. 14:23.000 --> 14:25.000 Where? 14:25.000 --> 14:27.000 There is. 14:27.000 --> 14:28.000 The left. 14:28.000 --> 14:30.000 Oh yeah, I see. 14:39.000 --> 14:42.000 Hey, thanks for the talk first. 14:42.000 --> 14:47.000 And I was wondering if podmen also handles the life cycle of the pod, 14:47.000 --> 14:52.000 meaning it checks for the lifeless probes and all the rest of the people that does. 14:52.000 --> 14:55.000 Yeah, this is a feature that we. 14:55.000 --> 14:58.000 The, the, yeah, the. 14:58.000 --> 15:04.000 The lifeiness probe, the health check are our feature that we support and can we can restart pod. 15:04.000 --> 15:07.000 Automatically if. 15:07.000 --> 15:11.000 If so, this is something that we support at the container level anyway. 15:11.000 --> 15:15.000 So this is not just the feature of pods, but it's a feature of containers. 15:15.000 --> 15:19.000 So it's you can define an L check and you can specify to restart the container. 15:19.000 --> 15:22.000 If the L check. 15:22.000 --> 15:25.000 Doesn't be as as expected. 15:30.000 --> 15:32.000 All right, thanks a lot. 15:32.000 --> 15:33.000 Any more questions? 15:33.000 --> 15:34.000 You can. 15:34.000 --> 15:36.000 Thank you.