1 00:00:02,120 --> 00:00:05,450 Now I will say that services like ECS, 2 00:00:05,450 --> 00:00:07,630 which we used in the deployment section 3 00:00:07,630 --> 00:00:09,670 can help us with these problems. 4 00:00:09,670 --> 00:00:12,200 And to some extent, we already used 5 00:00:12,200 --> 00:00:17,200 the help AWS ECS provides without even noticing it. 6 00:00:17,510 --> 00:00:22,430 For example, the ECS service will automatically check 7 00:00:22,430 --> 00:00:24,610 the health of our containers, 8 00:00:24,610 --> 00:00:27,070 so it will see if there are still up and running 9 00:00:27,070 --> 00:00:29,700 and it will automatically restart containers 10 00:00:29,700 --> 00:00:31,550 when they crash. 11 00:00:31,550 --> 00:00:34,370 And that is something we did see in action 12 00:00:34,370 --> 00:00:37,530 throughout the deployment section when our container, 13 00:00:37,530 --> 00:00:40,580 for example, failed to connect to a database 14 00:00:40,580 --> 00:00:44,483 and therefore was restarted and restarted and restarted. 15 00:00:45,450 --> 00:00:49,600 So AWS offers this to us, and when using ECS 16 00:00:49,600 --> 00:00:51,980 it's very easy to take advantage of that. 17 00:00:52,830 --> 00:00:56,390 And even though we didn't really utilize that a lot, 18 00:00:56,390 --> 00:01:01,020 AWS ECS, also would help us with auto scaling. 19 00:01:01,020 --> 00:01:04,730 So with dynamically, increasing and decreasing 20 00:01:04,730 --> 00:01:07,980 the number of container instances in the end, 21 00:01:07,980 --> 00:01:09,800 which are up and running 22 00:01:09,800 --> 00:01:13,980 to help us handle incoming traffic spikes, for example. 23 00:01:13,980 --> 00:01:17,410 And it then also has a tool called a load balancer, 24 00:01:17,410 --> 00:01:21,640 which we also did set up to have a unchanging IP address 25 00:01:21,640 --> 00:01:24,440 or a domain we could use for testing, 26 00:01:24,440 --> 00:01:26,980 which does actually not just give us this domain, 27 00:01:26,980 --> 00:01:28,290 which never changes, 28 00:01:28,290 --> 00:01:30,130 but which would also ensure 29 00:01:30,130 --> 00:01:32,090 that if you got incoming traffic, 30 00:01:32,090 --> 00:01:34,690 that traffic is split evenly 31 00:01:34,690 --> 00:01:37,420 across all running container instances, 32 00:01:37,420 --> 00:01:40,543 in case you have more than one running container instance. 33 00:01:41,770 --> 00:01:45,250 So services like AWS ECS 34 00:01:45,250 --> 00:01:49,200 already help us solve the problems I just outlined, 35 00:01:49,200 --> 00:01:51,300 which is another strong reason 36 00:01:51,300 --> 00:01:54,850 for turning towards such managed services 37 00:01:54,850 --> 00:01:58,440 instead of manually deploying everything. 38 00:01:58,440 --> 00:02:00,820 But of course, we still have one downside, 39 00:02:00,820 --> 00:02:05,680 one disadvantage when using a service like AWS ECS, 40 00:02:05,680 --> 00:02:09,870 we are basically locked into that service. 41 00:02:09,870 --> 00:02:14,100 If we use a specific cloud service by some cloud provider, 42 00:02:14,100 --> 00:02:16,870 no matter which cloud provider that is, 43 00:02:16,870 --> 00:02:20,203 we are kind of locked into that service. 44 00:02:21,120 --> 00:02:23,380 At the example of ECS, 45 00:02:23,380 --> 00:02:25,050 the problem which we have 46 00:02:25,050 --> 00:02:30,050 is that we need to configure everything as defined by AWS. 47 00:02:31,159 --> 00:02:36,100 AWS, for example, things and clusters, tasks and services, 48 00:02:36,100 --> 00:02:41,033 and we need to set the options, AWS lets us set. 49 00:02:41,870 --> 00:02:46,670 So whatever we set up there will then work for AWS ECS, 50 00:02:46,670 --> 00:02:51,090 and we can define it here in this visual user interface 51 00:02:51,090 --> 00:02:56,090 in the AWS console or all the work with a command line tool 52 00:02:56,110 --> 00:02:59,067 like the ecs-cli, which we haven't used, 53 00:02:59,067 --> 00:03:02,960 but we don't have to use that console if we don't want to, 54 00:03:02,960 --> 00:03:04,570 we can also use the command line 55 00:03:04,570 --> 00:03:07,270 to create all these resources. 56 00:03:07,270 --> 00:03:11,160 And we can also actually define configuration files 57 00:03:11,160 --> 00:03:14,110 where we basically write down the configuration 58 00:03:14,110 --> 00:03:18,320 we wanna have for our ECS cluster and tasks. 59 00:03:18,320 --> 00:03:23,320 And then we can use that AWS ECS-CLI to pick up 60 00:03:23,350 --> 00:03:25,470 on these configuration files 61 00:03:25,470 --> 00:03:29,130 and create the appropriate cloud resources based on that. 62 00:03:29,130 --> 00:03:31,330 So we don't have to use that UI, 63 00:03:31,330 --> 00:03:34,990 we can also use conflict files and therefore maybe speed up 64 00:03:34,990 --> 00:03:36,960 the entire deployment process, 65 00:03:36,960 --> 00:03:41,893 but we will always be locked into this specific AWS service, 66 00:03:42,820 --> 00:03:45,960 which means if we wanna switch to another service 67 00:03:45,960 --> 00:03:47,930 by another cloud provider, 68 00:03:47,930 --> 00:03:51,920 the conflict files we wrote for AWS ECS 69 00:03:51,920 --> 00:03:56,490 won't work there because this is AWS specific. 70 00:03:56,490 --> 00:04:01,280 The settings defined here will work with AWS ECS, 71 00:04:01,280 --> 00:04:02,740 but not with some service 72 00:04:02,740 --> 00:04:05,610 offered by Microsoft Azure, for example. 73 00:04:05,610 --> 00:04:09,980 So we're locking ourselves into AWS and AWS ECS 74 00:04:09,980 --> 00:04:11,460 if we start working with it 75 00:04:11,460 --> 00:04:14,720 and we need to learn the specifics of that service 76 00:04:14,720 --> 00:04:18,270 the specific terms and the configuration options we can set. 77 00:04:18,270 --> 00:04:20,710 And if we wanna switch to another provider, 78 00:04:20,710 --> 00:04:23,860 we have to learn the specifics of that provider 79 00:04:23,860 --> 00:04:26,210 and the service offered by that provider, 80 00:04:26,210 --> 00:04:30,130 and we manually have to translate the configuration 81 00:04:30,130 --> 00:04:35,130 we used for AWS ECS to that service by some other provider, 82 00:04:35,630 --> 00:04:37,073 if we wanna switch. 83 00:04:37,940 --> 00:04:41,180 So just knowing Docker won't be enough, 84 00:04:41,180 --> 00:04:43,540 we also have to learn about AWS ECS 85 00:04:44,940 --> 00:04:47,063 or whichever service we're using. 86 00:04:48,090 --> 00:04:50,900 Now, of course, this might be totally fine 87 00:04:50,900 --> 00:04:53,630 if you know that you wanna stick with AWS, 88 00:04:53,630 --> 00:04:55,170 you're happy with it, 89 00:04:55,170 --> 00:04:58,550 and once you learn how to work with ECS, 90 00:04:58,550 --> 00:05:01,270 of course, that's just a skill you have 91 00:05:01,270 --> 00:05:02,710 and you will be fine. 92 00:05:02,710 --> 00:05:06,000 So this might not be a huge problem necessarily, 93 00:05:06,000 --> 00:05:08,480 but of course it can be an issue, 94 00:05:08,480 --> 00:05:10,953 and that's where Kubernetes us can help us.