1 00:00:02,180 --> 00:00:04,090 So now, we're going to repeat 2 00:00:04,090 --> 00:00:06,740 the deployment example from before, 3 00:00:06,740 --> 00:00:09,770 but now not with AWS EC2, 4 00:00:09,770 --> 00:00:12,000 where we need to manage on our own, 5 00:00:12,000 --> 00:00:15,390 but with another service, which we can find by clicking 6 00:00:15,390 --> 00:00:17,280 on that service's dropdown here, 7 00:00:17,280 --> 00:00:19,220 and then searching for ECS, 8 00:00:19,220 --> 00:00:21,910 which is this Elastic Container Service. 9 00:00:21,910 --> 00:00:24,300 That's what we're going to use now. 10 00:00:24,300 --> 00:00:27,480 And now, I have one important disclaimer here. 11 00:00:27,480 --> 00:00:31,380 Where EC2, the way we used it at least, 12 00:00:31,380 --> 00:00:34,030 was covered by that free tier, 13 00:00:34,030 --> 00:00:37,060 the services we're using now and over the next lectures 14 00:00:37,060 --> 00:00:39,150 might not be covered. 15 00:00:39,150 --> 00:00:41,700 So if you wanna avoid any costs, 16 00:00:41,700 --> 00:00:45,130 you definitely should follow along passively here 17 00:00:45,130 --> 00:00:48,780 and/or check the AWS pricing pages 18 00:00:48,780 --> 00:00:52,430 to find out which costs you might incur. 19 00:00:52,430 --> 00:00:56,620 Now, we're only going to use the minimal set of features, 20 00:00:56,620 --> 00:00:59,390 so potential costs shouldn't be high, 21 00:00:59,390 --> 00:01:01,810 but I still wanna emphasize it here 22 00:01:01,810 --> 00:01:06,220 so that you don't blame me if your credit card gets charged. 23 00:01:06,220 --> 00:01:07,840 It's a cloud service provider. 24 00:01:07,840 --> 00:01:10,280 They earn money by selling services, 25 00:01:10,280 --> 00:01:13,030 and therefore, of course, not everything is free. 26 00:01:13,030 --> 00:01:14,830 That's just a disclaimer here. 27 00:01:14,830 --> 00:01:16,780 And just to also make this clear, 28 00:01:16,780 --> 00:01:18,660 we would have the same problem 29 00:01:18,660 --> 00:01:20,800 with basically any provider 30 00:01:20,800 --> 00:01:24,620 because ultimately some services will cost money. 31 00:01:24,620 --> 00:01:25,850 So with that out of the way, 32 00:01:25,850 --> 00:01:27,983 let's dive in and use that service though. 33 00:01:29,120 --> 00:01:32,620 And here, you should see some getting started screen, 34 00:01:32,620 --> 00:01:35,920 which might look slightly different after some time, 35 00:01:35,920 --> 00:01:38,650 but in the end, you should have some getting started screen 36 00:01:38,650 --> 00:01:40,820 with a get started button. 37 00:01:40,820 --> 00:01:44,720 Press that button, and you should see something like this. 38 00:01:44,720 --> 00:01:46,990 That's the wizard you will see 39 00:01:46,990 --> 00:01:51,560 when you launch your first container with ECS. 40 00:01:51,560 --> 00:01:54,150 Now, the container we are going to launch 41 00:01:54,150 --> 00:01:57,340 is that node-example-1 container, 42 00:01:57,340 --> 00:02:00,550 which I pushed to Docker Hub before 43 00:02:00,550 --> 00:02:02,450 in the last lectures. 44 00:02:02,450 --> 00:02:04,850 So if you skipped these lectures, 45 00:02:04,850 --> 00:02:08,930 make sure you take the example project you find attached, 46 00:02:08,930 --> 00:02:11,480 build an image from that project, 47 00:02:11,480 --> 00:02:14,850 and push that image to your Docker Hub account, 48 00:02:14,850 --> 00:02:16,310 and then use that image 49 00:02:16,310 --> 00:02:20,000 because this specific image on my Docker Hub account 50 00:02:20,000 --> 00:02:23,453 will not be available anymore when you're watching this. 51 00:02:24,410 --> 00:02:26,660 For me, right now, it is available though, 52 00:02:26,660 --> 00:02:29,470 and that's the image which we'll now deploy 53 00:02:29,470 --> 00:02:31,380 with help of ECS. 54 00:02:31,380 --> 00:02:33,587 So how does this work then? 55 00:02:33,587 --> 00:02:38,587 AWS ECS thinks in four categories, 56 00:02:38,610 --> 00:02:42,600 clusters, containers, tasks, and services. 57 00:02:42,600 --> 00:02:44,730 Now, it should be clear what containers are, 58 00:02:44,730 --> 00:02:46,850 and that's what we're going to start with. 59 00:02:46,850 --> 00:02:48,410 Initially in that wizard, 60 00:02:48,410 --> 00:02:51,090 we pick the container we wanna deploy. 61 00:02:51,090 --> 00:02:53,450 Now, there are some examples here, 62 00:02:53,450 --> 00:02:56,220 but we're going to go with our own custom container. 63 00:02:56,220 --> 00:02:58,050 So we should click on Configure here, 64 00:02:58,050 --> 00:03:01,020 and this opens up this side drawer 65 00:03:01,020 --> 00:03:04,093 where we can now define our own custom container. 66 00:03:04,940 --> 00:03:06,900 Now, we should give this container a name, 67 00:03:06,900 --> 00:03:09,890 and I'll choose node-demo here, 68 00:03:09,890 --> 00:03:12,320 and then we have to define the image 69 00:03:12,320 --> 00:03:14,730 on which this container should be based. 70 00:03:14,730 --> 00:03:16,800 And before we do that, 71 00:03:16,800 --> 00:03:20,840 I have one important note about this side drawer. 72 00:03:20,840 --> 00:03:23,800 In the end here, we just specify 73 00:03:23,800 --> 00:03:28,493 how ECS should later execute docker run. 74 00:03:29,500 --> 00:03:31,840 You will see plenty of options here, 75 00:03:31,840 --> 00:03:36,470 which you could also just plug into your docker run command. 76 00:03:36,470 --> 00:03:41,039 For example, --name for the container name. 77 00:03:41,039 --> 00:03:43,990 And the image is, of course, also the required 78 00:03:43,990 --> 00:03:47,073 for docker run, so we also need to define it here. 79 00:03:47,948 --> 00:03:50,940 In here, you should take your repository name. 80 00:03:50,940 --> 00:03:54,783 In my case, it's academind/node-example-1. 81 00:03:55,770 --> 00:03:58,510 And if your image is on Docker Hub, 82 00:03:58,510 --> 00:04:01,370 you just need to enter this repository name, 83 00:04:01,370 --> 00:04:05,690 and ECS will automatically look on Docker Hub. 84 00:04:05,690 --> 00:04:08,650 If you're using some other container registry, 85 00:04:08,650 --> 00:04:13,650 you need to have the full registry hosting domain here, 86 00:04:13,800 --> 00:04:16,230 and then the repository name. 87 00:04:16,230 --> 00:04:20,360 But for Docker Hub, just the repository name is enough. 88 00:04:20,360 --> 00:04:22,750 Now, if it should be a private repository, 89 00:04:22,750 --> 00:04:26,170 you can specify your login information here, 90 00:04:26,170 --> 00:04:29,380 but it's not a private repository here. 91 00:04:29,380 --> 00:04:32,220 You can leave the default for the memory limits, 92 00:04:32,220 --> 00:04:34,910 but of course we wanna map some ports. 93 00:04:34,910 --> 00:04:37,980 Again, in the end here, in this overlay, 94 00:04:37,980 --> 00:04:42,530 we just define how docker run should be executed later. 95 00:04:42,530 --> 00:04:46,970 And if you run it locally, you also might specify a name, 96 00:04:46,970 --> 00:04:51,720 but we then, for this project, which listens on port 80, 97 00:04:51,720 --> 00:04:54,900 would definitely also publish port 80 98 00:04:54,900 --> 00:04:58,430 on the host machine with the -p flag. 99 00:04:58,430 --> 00:05:01,850 And that's just what we define here with port mappings. 100 00:05:01,850 --> 00:05:04,400 We want to expose port 80. 101 00:05:04,400 --> 00:05:07,810 Now, you don't need to specify two numbers here 102 00:05:07,810 --> 00:05:11,570 because the container internal port will always be mapped 103 00:05:11,570 --> 00:05:14,310 to the same port outside of the container. 104 00:05:14,310 --> 00:05:16,210 So this is automatically mapping 105 00:05:16,210 --> 00:05:19,913 the container internal port 80 to the outside port 80. 106 00:05:21,230 --> 00:05:25,510 Now, you can also configure advanced container options. 107 00:05:25,510 --> 00:05:27,830 Now, we can ignore the healthcheck for now. 108 00:05:27,830 --> 00:05:30,810 That's some AWS-specific setting 109 00:05:30,810 --> 00:05:33,700 which allows AWS to run a specific command 110 00:05:33,700 --> 00:05:37,280 to check whether the container is running successfully. 111 00:05:37,280 --> 00:05:38,900 We can ignore this here. 112 00:05:38,900 --> 00:05:41,530 But for the environment, we can, for example, 113 00:05:41,530 --> 00:05:45,000 override the default entry point or command 114 00:05:45,000 --> 00:05:48,618 that should be executed for this container when it starts. 115 00:05:48,618 --> 00:05:52,440 Again, that's also something you can do with docker run. 116 00:05:52,440 --> 00:05:56,470 You saw that in the utility containers section, for example. 117 00:05:56,470 --> 00:05:58,030 You can run a container 118 00:05:58,030 --> 00:06:01,740 and specify your own command that should be executed 119 00:06:01,740 --> 00:06:05,540 instead of the command specified in the Dockerfile. 120 00:06:05,540 --> 00:06:07,370 Here, however, I wanna go 121 00:06:07,370 --> 00:06:09,780 with the Dockerfile internal command, 122 00:06:09,780 --> 00:06:12,763 so I will not specify my own command. 123 00:06:13,620 --> 00:06:16,333 You could also override the working directory. 124 00:06:17,330 --> 00:06:19,260 And, again, even though that is nothing 125 00:06:19,260 --> 00:06:21,230 we've done thus far in the course, 126 00:06:21,230 --> 00:06:25,730 you indeed would have an option on docker run to do that. 127 00:06:25,730 --> 00:06:30,730 If we run docker run --help, we see all available options. 128 00:06:30,780 --> 00:06:34,000 And there, you see there is a workdir option 129 00:06:34,000 --> 00:06:37,120 which would allow you to override the working directory 130 00:06:37,120 --> 00:06:40,530 for this container when it's being started. 131 00:06:40,530 --> 00:06:42,870 And that's what you can set up here. 132 00:06:42,870 --> 00:06:44,680 Now, I wanna do none of that. 133 00:06:44,680 --> 00:06:48,240 I just wanna explain what you're configuring here. 134 00:06:48,240 --> 00:06:51,350 Now, if your container would be using environment variables, 135 00:06:51,350 --> 00:06:53,870 you probably would want to pass them in 136 00:06:53,870 --> 00:06:55,890 when the container runs. 137 00:06:55,890 --> 00:06:59,903 And locally, we did that with docker run --env. 138 00:07:01,190 --> 00:07:02,730 Now here, we can define 139 00:07:02,730 --> 00:07:05,790 our environment variable key value pairs 140 00:07:05,790 --> 00:07:10,403 simply here in this part of that AWS console. 141 00:07:12,090 --> 00:07:14,420 Now, we can also define container timeouts, 142 00:07:14,420 --> 00:07:18,570 so basically telling AWS when it should stop trying 143 00:07:18,570 --> 00:07:20,280 to launch this container. 144 00:07:20,280 --> 00:07:22,380 We can set up network settings, 145 00:07:22,380 --> 00:07:23,790 and I'll come back to that later 146 00:07:23,790 --> 00:07:26,560 once we have more than one container being deployed. 147 00:07:26,560 --> 00:07:28,920 For the moment, we can ignore all of that. 148 00:07:28,920 --> 00:07:32,980 By default, our container will be reachable from the web. 149 00:07:32,980 --> 00:07:34,760 I can tell you that. 150 00:07:34,760 --> 00:07:37,030 Unless you disable networking, of course, 151 00:07:37,030 --> 00:07:38,480 which we don't wanna do here. 152 00:07:38,480 --> 00:07:40,550 We wanna keep it enabled. 153 00:07:40,550 --> 00:07:42,020 Under storage and logging, 154 00:07:42,020 --> 00:07:45,180 you can define mount points and volumes. 155 00:07:45,180 --> 00:07:48,290 Again, that's not something we use right now, 156 00:07:48,290 --> 00:07:51,290 but I will come back to that later as well. 157 00:07:51,290 --> 00:07:54,820 And, of course, that is the equivalent to the -v option 158 00:07:54,820 --> 00:07:56,043 on docker run. 159 00:07:57,710 --> 00:08:01,150 Then, you can configure logging if you want to. 160 00:08:01,150 --> 00:08:04,230 And, with Auto-configure CloudWatch Logs, 161 00:08:04,230 --> 00:08:09,160 you automatically use another AWS service called CloudWatch 162 00:08:09,160 --> 00:08:13,120 to manage and store the logs generated by this container, 163 00:08:13,120 --> 00:08:15,680 which could be useful, so I'll do that here, 164 00:08:15,680 --> 00:08:17,760 but you don't have to check it. 165 00:08:17,760 --> 00:08:19,730 And that is basically it. 166 00:08:19,730 --> 00:08:23,030 Now with that, we can click Update here, 167 00:08:23,030 --> 00:08:26,003 and now we configured our container 168 00:08:26,003 --> 00:08:28,560 that should be launched here. 169 00:08:28,560 --> 00:08:29,810 Now, in case you're wondering 170 00:08:29,810 --> 00:08:32,480 how you could launch multiple containers, 171 00:08:32,480 --> 00:08:33,520 I'll come back to that. 172 00:08:33,520 --> 00:08:34,352 For the moment, 173 00:08:34,352 --> 00:08:37,063 let's stick to this single container example. 174 00:08:38,940 --> 00:08:42,210 Now, in the next section, we can define our task 175 00:08:42,210 --> 00:08:43,470 for this container. 176 00:08:43,470 --> 00:08:46,640 And the task is basically, as it says here, 177 00:08:46,640 --> 00:08:48,930 the blueprint for your application. 178 00:08:48,930 --> 00:08:52,950 Here, you can tell AWS how it should launch your container. 179 00:08:52,950 --> 00:08:55,420 So not how it should execute docker run, 180 00:08:55,420 --> 00:08:58,070 but how the server on which it runs this 181 00:08:58,070 --> 00:09:00,680 should be configured, you could say. 182 00:09:00,680 --> 00:09:04,040 And therefore a task actually also can include 183 00:09:04,040 --> 00:09:07,320 more than one container, and you will also see that later 184 00:09:07,320 --> 00:09:11,020 when we add multiple containers to one and the same task. 185 00:09:11,020 --> 00:09:15,460 So you can think of the task as one remote server, 186 00:09:15,460 --> 00:09:19,860 one remote machine, which runs one or more containers. 187 00:09:19,860 --> 00:09:23,690 So a bit like the EC2 instances you saw before, 188 00:09:23,690 --> 00:09:25,680 which we managed ourself. 189 00:09:25,680 --> 00:09:27,870 Now, we're not managing it ourself, 190 00:09:27,870 --> 00:09:30,140 but we're still telling AWS 191 00:09:30,140 --> 00:09:32,300 how it should execute our containers, 192 00:09:32,300 --> 00:09:35,750 and which environment it should generally set up for them. 193 00:09:35,750 --> 00:09:37,810 And that's what a task is. 194 00:09:37,810 --> 00:09:41,440 And here, we're using FARGATE by default. 195 00:09:41,440 --> 00:09:44,490 That's a specific way of launching your container, 196 00:09:44,490 --> 00:09:48,940 which in the end launches it in a so-called serverless mode, 197 00:09:48,940 --> 00:09:53,930 which means AWS does not really create an EC2 instance 198 00:09:53,930 --> 00:09:55,790 on which it runs your container, 199 00:09:55,790 --> 00:09:59,830 but instead it stores your container and your run settings, 200 00:09:59,830 --> 00:10:01,860 and whenever there is a request 201 00:10:01,860 --> 00:10:05,460 that requires the container to do something, 202 00:10:05,460 --> 00:10:08,890 it starts the container up, handles that request, 203 00:10:08,890 --> 00:10:10,450 and then stops it again, 204 00:10:10,450 --> 00:10:12,880 which is, of course, very cost effective 205 00:10:12,880 --> 00:10:14,670 because you only pay 206 00:10:14,670 --> 00:10:17,420 for the time your container is executing 207 00:10:17,420 --> 00:10:20,893 and not for the time where it's just sitting around idle. 208 00:10:21,820 --> 00:10:23,910 So that's why we're using FARGATE here, 209 00:10:23,910 --> 00:10:27,310 this serverless container launching mode. 210 00:10:27,310 --> 00:10:30,090 And then, we can click Next. 211 00:10:30,090 --> 00:10:32,830 Now here, we define our service. 212 00:10:32,830 --> 00:10:37,480 And a service, in the end, now controls how this task, 213 00:10:37,480 --> 00:10:40,650 so this configured application and the container 214 00:10:40,650 --> 00:10:44,185 that belongs to it, should be executed. 215 00:10:44,185 --> 00:10:47,480 And it's here where you, for example, 216 00:10:47,480 --> 00:10:49,810 could add a load balancer, 217 00:10:49,810 --> 00:10:52,410 and it manages all the heavy lifting 218 00:10:52,410 --> 00:10:55,200 of redirecting the incoming requests, 219 00:10:55,200 --> 00:10:57,160 queue the up and running containers, 220 00:10:57,160 --> 00:11:00,040 and all of that behind the scenes. 221 00:11:00,040 --> 00:11:02,780 Now, I'm not going to use a load balancer here, 222 00:11:02,780 --> 00:11:05,400 but we will add one later in this module. 223 00:11:05,400 --> 00:11:09,170 To keep things simple for now, we'll start without one. 224 00:11:09,170 --> 00:11:13,280 So you could say every task is executed by a service. 225 00:11:13,280 --> 00:11:15,500 So you have one service per task. 226 00:11:15,500 --> 00:11:19,350 And these are just the terms in which AWS thinks. 227 00:11:19,350 --> 00:11:22,190 You will, of course, see all these things in action 228 00:11:22,190 --> 00:11:24,993 and working together over the next lectures. 229 00:11:25,840 --> 00:11:27,640 So let's click Next then. 230 00:11:27,640 --> 00:11:31,580 And the last thing we need to configure is our cluster. 231 00:11:31,580 --> 00:11:34,640 You could say that's the overall network 232 00:11:34,640 --> 00:11:36,900 in which our services run. 233 00:11:36,900 --> 00:11:38,570 Here, we have just one service 234 00:11:38,570 --> 00:11:40,590 with one task with one container, 235 00:11:40,590 --> 00:11:42,530 but if you had a multi-container app, 236 00:11:42,530 --> 00:11:46,140 you could group multiple containers in one cluster 237 00:11:46,140 --> 00:11:48,880 so that they all belong together logically, 238 00:11:48,880 --> 00:11:51,790 and they all can talk to each other. 239 00:11:51,790 --> 00:11:54,960 Now here, we automatically get this cluster network 240 00:11:54,960 --> 00:11:57,570 created for us, and therefore the only thing 241 00:11:57,570 --> 00:12:00,580 we need to do now is click Next. 242 00:12:00,580 --> 00:12:02,470 This gives us this Review page, 243 00:12:02,470 --> 00:12:04,370 where we can have a look whether everything 244 00:12:04,370 --> 00:12:05,900 is configured the way we want, 245 00:12:05,900 --> 00:12:08,530 and then I'm gonna click on Create. 246 00:12:08,530 --> 00:12:12,190 AWS will set up all the things it needs 247 00:12:12,190 --> 00:12:15,360 to launch that container when needed 248 00:12:15,360 --> 00:12:17,750 in the way you told it to do so. 249 00:12:17,750 --> 00:12:20,270 Now, this can take a couple of minutes here, 250 00:12:20,270 --> 00:12:24,003 and I'll be back once setting up all of that finished. 251 00:12:26,380 --> 00:12:30,140 So now, this finished, everything was set up, 252 00:12:30,140 --> 00:12:32,663 and we can now click on view service. 253 00:12:33,690 --> 00:12:36,310 Now, this page might look overwhelming. 254 00:12:36,310 --> 00:12:40,210 Here, you can see all the details of the environment 255 00:12:40,210 --> 00:12:42,867 and the configuration that was created by AWS. 256 00:12:43,970 --> 00:12:46,100 Now, for us, it's probably most interesting 257 00:12:46,100 --> 00:12:48,760 to see our running application. 258 00:12:48,760 --> 00:12:51,420 And you can see this by clicking on Tasks, 259 00:12:51,420 --> 00:12:54,680 and then click on that task which was created here. 260 00:12:54,680 --> 00:12:57,060 Important, on that task ID, 261 00:12:57,060 --> 00:13:01,130 not on the task definition, but on the task ID. 262 00:13:01,130 --> 00:13:04,140 And here, you'll find a public IP. 263 00:13:04,140 --> 00:13:07,920 Now, you could map your own domain to that public IP. 264 00:13:07,920 --> 00:13:11,340 I'll just paste this public IP into the browser, 265 00:13:11,340 --> 00:13:15,020 and I see my running application. 266 00:13:15,020 --> 00:13:19,900 And that's how we can now run it with help of AWS ECS. 267 00:13:19,900 --> 00:13:22,410 Now, this might still look overwhelming, 268 00:13:22,410 --> 00:13:23,930 but the huge advantage here 269 00:13:23,930 --> 00:13:28,500 is that we didn't start any custom servers or machines. 270 00:13:28,500 --> 00:13:31,870 We did not install anything, and we're not responsible 271 00:13:31,870 --> 00:13:34,320 for keeping anything up to date. 272 00:13:34,320 --> 00:13:38,700 We just configured how AWS should execute our containers, 273 00:13:38,700 --> 00:13:40,230 and that is it. 274 00:13:40,230 --> 00:13:42,430 There's nothing else to it. 275 00:13:42,430 --> 00:13:43,910 So now, let me also show you 276 00:13:43,910 --> 00:13:47,690 how we could update our container or our image 277 00:13:47,690 --> 00:13:49,100 on which this container is based 278 00:13:49,100 --> 00:13:52,880 and reflect those changes on this running cluster 279 00:13:52,880 --> 00:13:55,390 before we then thereafter also have a look 280 00:13:55,390 --> 00:13:57,593 at multi-container applications.