1 00:00:02,160 --> 00:00:03,969 Now to get our hands dirty, 2 00:00:03,969 --> 00:00:07,590 attached to find a very simple web application, 3 00:00:07,590 --> 00:00:10,790 a very simple Node.js application 4 00:00:10,790 --> 00:00:13,200 which just has two end points, 5 00:00:13,200 --> 00:00:17,660 a get request to /nothing and one to /error 6 00:00:17,660 --> 00:00:20,250 which will then crash this node application 7 00:00:20,250 --> 00:00:21,660 with this code line. 8 00:00:21,660 --> 00:00:23,130 I'll come back to that later. 9 00:00:23,130 --> 00:00:27,093 And let's deploy this to our kubernetes cluster. 10 00:00:28,350 --> 00:00:30,360 Now there's one super important thing 11 00:00:30,360 --> 00:00:33,060 which is easy to forget now 12 00:00:33,060 --> 00:00:35,870 with all that new kubernetes knowledge. 13 00:00:35,870 --> 00:00:38,389 We still need to use Docker 14 00:00:38,389 --> 00:00:41,820 to create an image based on this application 15 00:00:41,820 --> 00:00:45,190 because kubernetes is all about containers. 16 00:00:45,190 --> 00:00:48,820 And in order to run containers, it needs an image, right? 17 00:00:48,820 --> 00:00:50,620 That does not change. 18 00:00:50,620 --> 00:00:52,010 So while you'll learn about images 19 00:00:52,010 --> 00:00:54,250 and containers still applies. 20 00:00:54,250 --> 00:00:56,820 The only difference with kubernetes will be 21 00:00:56,820 --> 00:00:59,670 that you don't run your containers on your own anymore 22 00:00:59,670 --> 00:01:02,300 but we still need to create the images 23 00:01:02,300 --> 00:01:04,882 which we want to run as containers. 24 00:01:05,820 --> 00:01:09,760 So that's why I also added a Dockerfile to this project. 25 00:01:09,760 --> 00:01:13,330 A very simple Node.js app Dockerfile 26 00:01:13,330 --> 00:01:15,980 as we saw it many times throughout this course. 27 00:01:15,980 --> 00:01:18,490 Setting a base image and the working directory, 28 00:01:18,490 --> 00:01:21,960 installing dependencies, copying over the source code, 29 00:01:21,960 --> 00:01:23,630 exposing port 8080 30 00:01:23,630 --> 00:01:26,440 because that is the port we're listening on here. 31 00:01:26,440 --> 00:01:30,630 And then running this app.js file with the node executable 32 00:01:30,630 --> 00:01:32,313 when the container starts. 33 00:01:33,210 --> 00:01:36,380 Now, we could absolutely run this application locally 34 00:01:36,380 --> 00:01:41,380 but of course, I now want to run this on my cluster. 35 00:01:41,390 --> 00:01:45,070 And for this I'll open up the terminal here in my IDE. 36 00:01:45,070 --> 00:01:48,920 And first of all, build this image with docker build 37 00:01:48,920 --> 00:01:53,090 and give it a name of maybe kub-first-app. 38 00:01:56,300 --> 00:01:57,350 That's my name. 39 00:01:57,350 --> 00:02:00,060 And of course you can pick any name you want. 40 00:02:00,060 --> 00:02:02,390 This will now build this as an image 41 00:02:02,390 --> 00:02:03,930 and for me it's super fast 42 00:02:03,930 --> 00:02:06,970 because I already built it off screen before. 43 00:02:06,970 --> 00:02:08,520 But there's nothing fancy about it, 44 00:02:08,520 --> 00:02:11,430 we just built this image as we did it multiple times 45 00:02:11,430 --> 00:02:12,530 throughout the course. 46 00:02:14,300 --> 00:02:16,460 Now, once the image is built, 47 00:02:16,460 --> 00:02:20,090 we kind of wanna send it to the Kubernetes cluster 48 00:02:20,090 --> 00:02:24,300 but of course not like this, but instead as part of a pod 49 00:02:24,300 --> 00:02:28,680 or actually as we learned as part of a deployment, 50 00:02:28,680 --> 00:02:30,257 which will then create the pod 51 00:02:30,257 --> 00:02:34,333 and they offered a container running into pod for us, 52 00:02:34,333 --> 00:02:37,023 and it will also manage it for us. 53 00:02:37,920 --> 00:02:39,380 And to do that, 54 00:02:39,380 --> 00:02:43,130 to send this instruction to the Kubernetes cluster, 55 00:02:43,130 --> 00:02:45,380 we first of all need to ensure 56 00:02:45,380 --> 00:02:48,040 that this cluster is up and running. 57 00:02:48,040 --> 00:02:50,320 Now, if you followed along in the last lectures, 58 00:02:50,320 --> 00:02:51,700 that should be the case, 59 00:02:51,700 --> 00:02:54,320 but you can run minikube status 60 00:02:54,320 --> 00:02:56,460 to check whether it is up and running. 61 00:02:56,460 --> 00:02:58,440 And you should see a bunch of running here, 62 00:02:58,440 --> 00:03:01,470 which means this cluster is up and running. 63 00:03:01,470 --> 00:03:02,540 If it's not, 64 00:03:02,540 --> 00:03:07,100 you can restart it with minikube start --driver=docker 65 00:03:07,100 --> 00:03:11,140 and then set the driver to virtual box if you installed that 66 00:03:11,140 --> 00:03:13,470 or give docker a try, 67 00:03:13,470 --> 00:03:16,640 which is actually pretty inception like, 68 00:03:16,640 --> 00:03:19,000 because we would then use docker 69 00:03:19,000 --> 00:03:22,210 to create a virtual environment, 70 00:03:22,210 --> 00:03:26,420 a containerized environment in this case for the cluster, 71 00:03:26,420 --> 00:03:28,961 which we simulate on our local machine 72 00:03:28,961 --> 00:03:33,961 to then test Kubernetes which again is all about containers. 73 00:03:34,200 --> 00:03:38,580 Boom, head explosion, but it should also work as a driver 74 00:03:38,580 --> 00:03:41,080 so that is a virtual cluster, 75 00:03:41,080 --> 00:03:42,770 this cluster in a virtual machine 76 00:03:42,770 --> 00:03:46,320 is created with help of docker instead of virtual box. 77 00:03:46,320 --> 00:03:47,980 And on windows 10 pro, 78 00:03:47,980 --> 00:03:50,563 you should also be able to use hyper-V. 79 00:03:51,970 --> 00:03:54,370 Now I already got the cluster up and running. 80 00:03:54,370 --> 00:03:57,630 So with that, we can now send our image 81 00:03:57,630 --> 00:04:00,970 or send the instruction to create a deployment 82 00:04:00,970 --> 00:04:02,555 to that cluster. 83 00:04:02,555 --> 00:04:05,547 And we do this with this kubectl, this kubectl command. 84 00:04:08,010 --> 00:04:10,690 Remember that we also installed this tool 85 00:04:10,690 --> 00:04:12,600 over the last lectures. 86 00:04:12,600 --> 00:04:13,760 This is the command 87 00:04:13,760 --> 00:04:17,149 which you always execute on your local machine. 88 00:04:17,149 --> 00:04:20,260 So this will not change once we move to the cloud. 89 00:04:20,260 --> 00:04:23,080 This always executes on the local machine 90 00:04:23,080 --> 00:04:27,270 because this is the command we use to send instructions 91 00:04:27,270 --> 00:04:29,390 to our kubernetes cluster. 92 00:04:29,390 --> 00:04:32,580 So to master node and to control plane 93 00:04:32,580 --> 00:04:34,453 in that cluster so to say. 94 00:04:35,390 --> 00:04:36,970 So this is the command we need. 95 00:04:36,970 --> 00:04:39,930 And now this actually supports a broad variety 96 00:04:39,930 --> 00:04:42,190 of things we can do. 97 00:04:42,190 --> 00:04:45,698 If you just type kubectl help, 98 00:04:45,698 --> 00:04:50,350 you get a long list of commands you can execute. 99 00:04:50,350 --> 00:04:53,870 So there's a lot you can do, and this can be overwhelming 100 00:04:53,870 --> 00:04:58,870 but we actually only need a fraction of these things here. 101 00:04:59,130 --> 00:05:00,720 Now, of course, you can always dive 102 00:05:00,720 --> 00:05:03,910 in the official kubernetes docs if you wanna learn more 103 00:05:03,910 --> 00:05:06,870 but I will also show you the most important things 104 00:05:06,870 --> 00:05:09,473 throughout this and the next modules. 105 00:05:11,860 --> 00:05:16,190 So here we wanna use the kubectl command 106 00:05:16,190 --> 00:05:19,570 to create a new deployment object. 107 00:05:19,570 --> 00:05:22,640 Because as I mentioned, you work with these objects, 108 00:05:22,640 --> 00:05:25,710 which are then picked up by the Kubernetes cluster 109 00:05:25,710 --> 00:05:27,380 to then do something. 110 00:05:27,380 --> 00:05:31,660 And you create objects with the kubectl create command. 111 00:05:31,660 --> 00:05:33,700 Now, if you just hit enter like this, 112 00:05:33,700 --> 00:05:36,280 you actually get help for this command. 113 00:05:36,280 --> 00:05:39,240 And you can see what you can create there, 114 00:05:39,240 --> 00:05:41,290 basically these things. 115 00:05:41,290 --> 00:05:44,940 And for the moment we need to create a deployment object. 116 00:05:44,940 --> 00:05:48,690 And this is one of the most common objects you will create. 117 00:05:48,690 --> 00:05:52,640 As a side note at the moment, by using this create command, 118 00:05:52,640 --> 00:05:57,640 we are using this imperative approach of creating objects. 119 00:05:58,270 --> 00:06:02,550 Later, I will also show you the declarative approach. 120 00:06:02,550 --> 00:06:04,370 But for the moment we run create 121 00:06:04,370 --> 00:06:06,380 and we run create deployment, 122 00:06:06,380 --> 00:06:08,860 to create a new deployment object. 123 00:06:08,860 --> 00:06:11,730 And this object is then also automatically sent 124 00:06:11,730 --> 00:06:13,280 to the Kubernetes cluster. 125 00:06:13,280 --> 00:06:17,100 So we don't need to run a separate command for this. 126 00:06:17,100 --> 00:06:18,940 Now this deployment should get a name 127 00:06:18,940 --> 00:06:20,650 and we could name it first-app, 128 00:06:20,650 --> 00:06:23,550 but of course, this name is totally up to you. 129 00:06:23,550 --> 00:06:27,390 And then we need to specify one important thing here. 130 00:06:27,390 --> 00:06:30,780 We need to pass one important option. 131 00:06:30,780 --> 00:06:35,660 And that's the --image option, where we wanna specify 132 00:06:35,660 --> 00:06:39,073 which image should be used for the container 133 00:06:39,073 --> 00:06:42,750 of the pod created by this deployment. 134 00:06:42,750 --> 00:06:45,670 And this is --image=. 135 00:06:45,670 --> 00:06:49,340 Now here, you could of course think that you simply use 136 00:06:49,340 --> 00:06:53,280 that local image we just created, kub-first-app. 137 00:06:53,280 --> 00:06:56,100 That's the image I just built locally. 138 00:06:56,100 --> 00:06:58,330 But this actually won't work 139 00:06:58,330 --> 00:07:00,350 as you will see if you hit enter. 140 00:07:00,350 --> 00:07:03,040 You actually get a success message here. 141 00:07:03,040 --> 00:07:08,040 But if we now run another command, kubectl get deployments, 142 00:07:08,270 --> 00:07:12,050 we can see how many deployments we have in our cluster 143 00:07:12,050 --> 00:07:13,620 to which we connected here. 144 00:07:13,620 --> 00:07:17,280 By the way, kubectl was configured by minikube 145 00:07:17,280 --> 00:07:19,740 to automatically connect to minikube 146 00:07:19,740 --> 00:07:21,303 so we don't need to tell the command 147 00:07:21,303 --> 00:07:24,550 that we wanna interact with minikube here. 148 00:07:24,550 --> 00:07:26,180 So if I now run gets deployments, 149 00:07:26,180 --> 00:07:29,542 I see all the deployments in my minikube demo cluster 150 00:07:29,542 --> 00:07:33,090 and we see the deployment we just made. 151 00:07:33,090 --> 00:07:35,130 But we also see this. 152 00:07:35,130 --> 00:07:39,670 We have one deployment and zero of them are ready. 153 00:07:39,670 --> 00:07:43,760 Which means one deployment of one deployment failed 154 00:07:43,760 --> 00:07:47,653 which is not the best ratio to put it like that. 155 00:07:48,780 --> 00:07:53,040 Now to get or insights, we can run kubectl get pods 156 00:07:53,040 --> 00:07:57,630 and we'll now see all it's created by our deployments. 157 00:07:57,630 --> 00:08:01,750 And we see one part created by the first-app deployment 158 00:08:01,750 --> 00:08:05,090 but we see that here, one pod should be created. 159 00:08:05,090 --> 00:08:07,560 This is the target states so to say, 160 00:08:07,560 --> 00:08:10,740 but we don't have one pod created. 161 00:08:10,740 --> 00:08:15,720 And here we see the reason it's an ImagePull error 162 00:08:15,720 --> 00:08:19,200 and it should be pretty clear what's causing this. 163 00:08:19,200 --> 00:08:20,858 When we created deployment 164 00:08:20,858 --> 00:08:23,480 with this create deployment command 165 00:08:23,480 --> 00:08:25,930 the image which we specify here 166 00:08:25,930 --> 00:08:28,620 is not taken from the local machine 167 00:08:28,620 --> 00:08:31,480 and sent to the virtual machine. 168 00:08:31,480 --> 00:08:35,261 Instead, this command is sent to the Kubernetes cluster 169 00:08:35,261 --> 00:08:39,159 and then the cluster will look for this image. 170 00:08:39,159 --> 00:08:41,734 And since the cluster runs in a virtual machine, 171 00:08:41,734 --> 00:08:44,220 which I guess technically of course, 172 00:08:44,220 --> 00:08:46,090 is running on our local machine 173 00:08:46,090 --> 00:08:49,840 but which actually acts like a totally separate machine. 174 00:08:49,840 --> 00:08:53,550 This image, which only exists on our local machine, 175 00:08:53,550 --> 00:08:57,223 but not in the Kubernetes cluster of course isn't found. 176 00:08:58,350 --> 00:09:02,220 So instead of sending or specifying a local image here, 177 00:09:02,220 --> 00:09:03,960 we should specify an image 178 00:09:03,960 --> 00:09:08,700 which is part of a image registry like Docker Hub. 179 00:09:08,700 --> 00:09:11,690 So they affirmed let's for the moment, 180 00:09:11,690 --> 00:09:15,190 delete this first-app deployment. 181 00:09:15,190 --> 00:09:18,533 So run Kubectl delete deployment first-app, 182 00:09:18,533 --> 00:09:20,673 so that we remove that again. 183 00:09:21,800 --> 00:09:25,130 Thereafter you'll have no pods anymore. 184 00:09:25,130 --> 00:09:29,130 And let's now push this image from our local machine 185 00:09:29,130 --> 00:09:30,243 to Docker Hub. 186 00:09:32,120 --> 00:09:35,000 For that on your Docker Hub account, 187 00:09:35,000 --> 00:09:39,610 create a new repository and I'll name mine kub-first-app. 188 00:09:39,610 --> 00:09:41,230 And therefore it can later be found 189 00:09:41,230 --> 00:09:45,300 under your account named slash this name, 190 00:09:45,300 --> 00:09:48,880 create it, and then you learned how that works. 191 00:09:48,880 --> 00:09:53,880 Let's re-tag our local kub-first-app image 192 00:09:54,300 --> 00:09:57,730 as your Docker Hub account named slash 193 00:09:57,730 --> 00:10:00,220 and then your repository name. 194 00:10:00,220 --> 00:10:03,180 So in my case, I tag it like this 195 00:10:03,180 --> 00:10:06,100 and thereafter we can run docker push 196 00:10:06,100 --> 00:10:09,120 and then this repository image name. 197 00:10:09,120 --> 00:10:12,840 And now it does this being pushed to Docker hub, 198 00:10:12,840 --> 00:10:15,260 and that means that now it can be accessed 199 00:10:15,260 --> 00:10:19,943 from anywhere in the world, including our cluster machine. 200 00:10:21,210 --> 00:10:23,840 So with that if we now run this again, 201 00:10:23,840 --> 00:10:26,060 the create deployment command 202 00:10:26,060 --> 00:10:28,930 but now we point at this remote image. 203 00:10:28,930 --> 00:10:30,920 So this image on Docker Hub, 204 00:10:30,920 --> 00:10:34,040 and hence we adjust the image name here. 205 00:10:34,040 --> 00:10:36,120 If we tried this now, 206 00:10:36,120 --> 00:10:39,630 you will see that if you now get your deployments, 207 00:10:39,630 --> 00:10:43,793 you will eventually have one over one ready deployments. 208 00:10:44,790 --> 00:10:47,260 Initially, whilst it's starting up, 209 00:10:47,260 --> 00:10:49,670 you could still see zero over one here, 210 00:10:49,670 --> 00:10:53,080 but eventually you will have one over one here. 211 00:10:53,080 --> 00:10:56,200 And therefore, if you run get pods, 212 00:10:56,200 --> 00:10:59,400 you will also have one over one ready pod here 213 00:10:59,400 --> 00:11:01,470 with a status of running. 214 00:11:01,470 --> 00:11:03,443 So this is looking good now, 215 00:11:03,443 --> 00:11:06,820 our application is up and running. 216 00:11:06,820 --> 00:11:10,380 Of course at the moment, we can't really reach it though. 217 00:11:10,380 --> 00:11:12,680 So that's something we're going to work on, 218 00:11:12,680 --> 00:11:17,230 but to see at least something, we can run minikube dashboard 219 00:11:17,230 --> 00:11:20,030 which you might remember is this web dashboard 220 00:11:20,030 --> 00:11:21,530 which we can open up. 221 00:11:21,530 --> 00:11:25,020 This now is a process which needs to stay up and running 222 00:11:25,020 --> 00:11:28,000 at least as long as you wanna visit that dashboard. 223 00:11:28,000 --> 00:11:29,860 And it should open up the dashboard 224 00:11:29,860 --> 00:11:32,370 and the new browser window automatically 225 00:11:32,370 --> 00:11:35,280 if not, simply use that link, which you get here. 226 00:11:35,280 --> 00:11:38,650 And this is now our cluster. 227 00:11:38,650 --> 00:11:40,850 And here you should see your deployment 228 00:11:40,850 --> 00:11:42,350 the first-app deployment 229 00:11:42,350 --> 00:11:45,500 and also all the pods which are up and running. 230 00:11:45,500 --> 00:11:48,270 You'll also see that this pod has a label 231 00:11:48,270 --> 00:11:51,450 which is basically some metadata attached to it, 232 00:11:51,450 --> 00:11:54,230 which identifies the application 233 00:11:54,230 --> 00:11:56,770 and the deployment that belongs to. 234 00:11:56,770 --> 00:11:59,270 And you can also look into these things 235 00:11:59,270 --> 00:12:02,980 and you'll see more information about the deployment 236 00:12:02,980 --> 00:12:05,250 it's status and so on. 237 00:12:05,250 --> 00:12:07,078 And if you look into your pod, 238 00:12:07,078 --> 00:12:09,330 you also can learn more about that. 239 00:12:09,330 --> 00:12:12,720 For example, you can also find an IP address here 240 00:12:12,720 --> 00:12:16,090 but this will only be the internal IP address 241 00:12:16,090 --> 00:12:17,680 inside of the cluster. 242 00:12:17,680 --> 00:12:22,600 So if you try using that in your browser to reach that pod, 243 00:12:22,600 --> 00:12:23,840 this won't work, 244 00:12:23,840 --> 00:12:28,230 even if you use the port exposed by the container. 245 00:12:28,230 --> 00:12:29,820 So for the moment this won't work 246 00:12:29,820 --> 00:12:31,880 but we'll see how we can reach this pod 247 00:12:31,880 --> 00:12:33,683 and this container later. 248 00:12:34,840 --> 00:12:39,330 So that is all working, now I wanna take a brief look 249 00:12:39,330 --> 00:12:41,610 behind the scenes of what happened here 250 00:12:41,610 --> 00:12:43,110 before we then ensure 251 00:12:43,110 --> 00:12:46,160 that we can actually see and visit this webpage, 252 00:12:46,160 --> 00:12:48,015 which was created by the container, 253 00:12:48,015 --> 00:12:51,083 which has part off that pod we deployed.