1 00:00:02,130 --> 00:00:04,600 Now in order to get our hands dirty, 2 00:00:04,600 --> 00:00:06,800 and to start working with Kubernetes, 3 00:00:06,800 --> 00:00:11,700 there is one really, really, really important thing you got 4 00:00:11,700 --> 00:00:14,100 to understand and keep in mind. 5 00:00:14,100 --> 00:00:18,360 And that is the split between work you as a developer 6 00:00:18,360 --> 00:00:20,960 and administrator have to do 7 00:00:20,960 --> 00:00:24,180 and the work Kubernetes will do for you. 8 00:00:24,180 --> 00:00:26,730 Because there is an important difference 9 00:00:26,730 --> 00:00:31,190 and it's crucial that you understand what Kubernetes does 10 00:00:31,190 --> 00:00:33,170 and what it does not. 11 00:00:33,170 --> 00:00:35,090 And there is one super important thing 12 00:00:35,090 --> 00:00:36,980 which Kubernetes will not do 13 00:00:36,980 --> 00:00:39,220 which you instead have to do. 14 00:00:39,220 --> 00:00:43,150 And that is the cluster and node creation. 15 00:00:43,150 --> 00:00:45,120 Now, you may be expected 16 00:00:45,120 --> 00:00:48,890 that Kubernetes is this all in one tool 17 00:00:48,890 --> 00:00:52,840 where you define your final architecture, 18 00:00:52,840 --> 00:00:55,790 so where you define your cluster, 19 00:00:55,790 --> 00:00:59,290 and your master node and worker node configurations, 20 00:00:59,290 --> 00:01:03,260 and Kubernetes creates all these resources for you. 21 00:01:03,260 --> 00:01:08,260 But I have to tell you that is not what Kubernetes does. 22 00:01:08,370 --> 00:01:10,060 And I actually mentioned this 23 00:01:10,060 --> 00:01:12,870 in the last course section already. 24 00:01:12,870 --> 00:01:16,460 Kubernetes helps you with managing these pods 25 00:01:16,460 --> 00:01:18,170 with monitoring the pods, 26 00:01:18,170 --> 00:01:20,400 monitoring the containers in the pods, 27 00:01:20,400 --> 00:01:24,210 replacing failing pods and scaling pods. 28 00:01:24,210 --> 00:01:27,670 It's really just about orchestrating these containers 29 00:01:27,670 --> 00:01:29,360 which are inside of the pods, 30 00:01:29,360 --> 00:01:32,510 and of moving them around between worker nodes 31 00:01:32,510 --> 00:01:35,580 and ensuring that everything is up and running 32 00:01:35,580 --> 00:01:38,210 and reachable by your end users, 33 00:01:38,210 --> 00:01:40,870 this is what Kubernetes does. 34 00:01:40,870 --> 00:01:42,630 So put in other words, 35 00:01:42,630 --> 00:01:45,140 Kubernetes is responsible 36 00:01:45,140 --> 00:01:49,380 for managing your deployed application for ensuring 37 00:01:49,380 --> 00:01:52,600 that your application runs the way it should run, 38 00:01:52,600 --> 00:01:54,660 that your containers are running, 39 00:01:54,660 --> 00:01:56,660 everything related to that. 40 00:01:56,660 --> 00:01:59,220 Kubernetes will not take care 41 00:01:59,220 --> 00:02:03,080 about the infrastructure your application needs. 42 00:02:03,080 --> 00:02:05,060 It doesn't know anything about that 43 00:02:05,060 --> 00:02:09,449 and it's not a cloud infrastructure creation tool 44 00:02:09,449 --> 00:02:11,510 or anything like that. 45 00:02:11,510 --> 00:02:14,670 As I also mentioned in the last section already, 46 00:02:14,670 --> 00:02:17,860 Kubernetes is not a cloud service provider 47 00:02:17,860 --> 00:02:20,150 and also not a specific service 48 00:02:20,150 --> 00:02:22,500 by one cloud service provider. 49 00:02:22,500 --> 00:02:24,880 Instead, it's a framework and a collection 50 00:02:24,880 --> 00:02:29,640 of concepts and tools which allows you to set up deployments 51 00:02:29,640 --> 00:02:33,110 for your containerized applications. 52 00:02:33,110 --> 00:02:34,780 But therefore, Kubernetes, 53 00:02:34,780 --> 00:02:37,290 of course, doesn't know anything 54 00:02:37,290 --> 00:02:39,940 about the concrete remote machines 55 00:02:39,940 --> 00:02:41,760 to concrete virtual instances 56 00:02:41,760 --> 00:02:44,290 or computers you wanna work with, 57 00:02:44,290 --> 00:02:45,790 it doesn't know them. 58 00:02:45,790 --> 00:02:49,520 So you have to create these machines in advance 59 00:02:49,520 --> 00:02:53,740 and you have to install the Kubernetes API server, 60 00:02:53,740 --> 00:02:57,450 the kubelets and all these different services I mentioned 61 00:02:57,450 --> 00:03:00,897 in the last module on these master 62 00:03:00,897 --> 00:03:03,830 and worker nodes which you have to create. 63 00:03:03,830 --> 00:03:07,100 So you have to do all that instance creation, 64 00:03:07,100 --> 00:03:08,920 that instance management, 65 00:03:08,920 --> 00:03:12,870 and also the installation of all that software. 66 00:03:12,870 --> 00:03:15,270 And if you're running your own data center, 67 00:03:15,270 --> 00:03:16,840 that of course, means that you need 68 00:03:16,840 --> 00:03:19,250 to manage that entire data center 69 00:03:19,250 --> 00:03:22,300 and prepare everything therefore, Kubernetes. 70 00:03:22,300 --> 00:03:24,600 If you are using a cloud provider, 71 00:03:24,600 --> 00:03:26,970 it also means that you need to create, 72 00:03:26,970 --> 00:03:29,010 for example, with AWS, 73 00:03:29,010 --> 00:03:31,270 all the EC2 instances you need 74 00:03:31,270 --> 00:03:33,410 for the worker and master nodes 75 00:03:33,410 --> 00:03:35,400 and that you then also might need 76 00:03:35,400 --> 00:03:39,470 to create additional resources like a load balancer, 77 00:03:39,470 --> 00:03:43,621 or a file systems like the Elastic File System, 78 00:03:43,621 --> 00:03:48,621 AWS offers if your dockerized application would need that. 79 00:03:48,690 --> 00:03:50,370 And that, of course, brings us back 80 00:03:50,370 --> 00:03:54,300 to the manual deployment part which is exactly 81 00:03:54,300 --> 00:03:56,223 what we didn't want to do. 82 00:03:57,100 --> 00:04:01,310 Because with that we are facing various disadvantages. 83 00:04:01,310 --> 00:04:03,620 We have to keep these instances, 84 00:04:03,620 --> 00:04:05,290 these machines updated, 85 00:04:05,290 --> 00:04:07,250 keep the operating system updated, 86 00:04:07,250 --> 00:04:10,980 and manage the entire network security group stuff. 87 00:04:10,980 --> 00:04:14,780 So we have to ensure the security of these instances 88 00:04:14,780 --> 00:04:17,079 and of all these machines. 89 00:04:17,079 --> 00:04:18,350 And in addition, 90 00:04:18,350 --> 00:04:20,920 as mentioned in the last course module, 91 00:04:20,920 --> 00:04:24,270 we might face even more problems when it then comes 92 00:04:24,270 --> 00:04:27,483 to replacing failing containers and so on. 93 00:04:28,540 --> 00:04:31,750 Well, the good news is these problems are exactly 94 00:04:31,750 --> 00:04:33,850 what Kubernetes does takes care of. 95 00:04:33,850 --> 00:04:36,330 I mentioned this in the last core section 96 00:04:36,330 --> 00:04:38,640 and this is still true. 97 00:04:38,640 --> 00:04:40,340 Now when it comes to the security 98 00:04:40,340 --> 00:04:42,550 and managing these machines, 99 00:04:42,550 --> 00:04:45,840 these instances and managing the operating system on therein 100 00:04:45,840 --> 00:04:47,960 and keeping everything up to date, 101 00:04:47,960 --> 00:04:51,120 that is something Kubernetes can't help you with 102 00:04:51,120 --> 00:04:54,190 because it's not a Server Administration tool, 103 00:04:54,190 --> 00:04:57,500 it's not a cloud provider and it's not a managed service 104 00:04:57,500 --> 00:05:00,020 that would take that away from you. 105 00:05:00,020 --> 00:05:04,550 It really just helps you with what happens on these servers. 106 00:05:04,550 --> 00:05:08,930 So Kubernetes will actually help you create the pods 107 00:05:08,930 --> 00:05:11,230 and run the containers in the pods, 108 00:05:11,230 --> 00:05:13,800 it will help you monitor and replace them 109 00:05:13,800 --> 00:05:15,960 and it will help you with scaling. 110 00:05:15,960 --> 00:05:19,210 In the end, you could say Kubernetes will simply utilize 111 00:05:19,210 --> 00:05:21,110 the resources you create, 112 00:05:21,110 --> 00:05:23,430 so it will utilize the worker nodes, 113 00:05:23,430 --> 00:05:25,820 these machines you are creating, 114 00:05:25,820 --> 00:05:28,980 and it will then help you achieve your deployment goals 115 00:05:28,980 --> 00:05:30,440 for the containers. 116 00:05:30,440 --> 00:05:33,760 It will not create the resources for you though, 117 00:05:33,760 --> 00:05:36,510 that is something you still have to do. 118 00:05:36,510 --> 00:05:39,620 Now, if you are an experienced server administrator, 119 00:05:39,620 --> 00:05:41,690 you can of course do all of that. 120 00:05:41,690 --> 00:05:45,650 If you're not there are additional tools like Kubermatic 121 00:05:45,650 --> 00:05:47,730 which is built for Kubernetes, 122 00:05:47,730 --> 00:05:50,850 but not part of Kubernetes which can help you 123 00:05:50,850 --> 00:05:54,430 with creating all the infrastructure you need 124 00:05:54,430 --> 00:05:57,250 for your Kubernetes managed to deployment 125 00:05:57,250 --> 00:06:00,400 and therefore, this could be an additional tool 126 00:06:00,400 --> 00:06:03,070 which helps you create these remote machines 127 00:06:03,070 --> 00:06:04,480 and these instances, 128 00:06:04,480 --> 00:06:06,850 for example, on AWS. 129 00:06:06,850 --> 00:06:08,580 So that's a tool which can help you 130 00:06:08,580 --> 00:06:11,440 if you don't wanna create everything manually there. 131 00:06:11,440 --> 00:06:13,940 So that could be a step in the right direction. 132 00:06:13,940 --> 00:06:16,630 It's not part of Kubernetes itself, though, 133 00:06:16,630 --> 00:06:18,620 because Kubernetes is really just 134 00:06:18,620 --> 00:06:20,510 about managing these containers 135 00:06:20,510 --> 00:06:23,970 and keeping your application up and running and scaling it. 136 00:06:23,970 --> 00:06:26,743 It's really important to understand that difference. 137 00:06:27,680 --> 00:06:31,070 In addition, certain cloud providers like again, 138 00:06:31,070 --> 00:06:36,030 AWS do have dedicated services like in AWS, 139 00:06:36,030 --> 00:06:39,970 EKS, Elastic Kubernetes Service which actually helps you 140 00:06:39,970 --> 00:06:41,810 with that as well. 141 00:06:41,810 --> 00:06:43,970 That is a service which will allow you 142 00:06:43,970 --> 00:06:46,910 to bring your own Kubernetes configuration 143 00:06:46,910 --> 00:06:48,410 so that you don't have 144 00:06:48,410 --> 00:06:52,720 to use the AWS specific configuration as you did have 145 00:06:52,720 --> 00:06:56,600 to use with ECS but EKS. 146 00:06:56,600 --> 00:06:59,590 So this Elastic Kubernetes Service will 147 00:06:59,590 --> 00:07:02,510 then also set up all the resources needed 148 00:07:02,510 --> 00:07:05,230 by your Kubernetes configuration. 149 00:07:05,230 --> 00:07:07,420 So it's services like this 150 00:07:07,420 --> 00:07:10,490 or tools like Kubermatic which I just showed you, 151 00:07:10,490 --> 00:07:13,690 which can help you complete the picture, 152 00:07:13,690 --> 00:07:15,990 and which can therefore help you getting rid of 153 00:07:15,990 --> 00:07:20,120 that right side of this slide of the steps you have 154 00:07:20,120 --> 00:07:21,073 to take care of. 155 00:07:22,050 --> 00:07:23,940 I just want you to be aware of that 156 00:07:23,940 --> 00:07:25,750 because it's easy to think 157 00:07:25,750 --> 00:07:28,780 that Kubernetes will also create all these machines 158 00:07:28,780 --> 00:07:29,613 which it needs. 159 00:07:29,613 --> 00:07:32,453 And that Kubernetes will install the kubelets 160 00:07:32,453 --> 00:07:35,490 and the API server and the scheduler 161 00:07:35,490 --> 00:07:37,230 and all the things we talked 162 00:07:37,230 --> 00:07:41,910 about when we analyzed the master and worker nodes. 163 00:07:41,910 --> 00:07:43,840 And it's really important to understand 164 00:07:43,840 --> 00:07:45,730 that Kubernetes will not do that. 165 00:07:45,730 --> 00:07:47,770 It will not create any machines 166 00:07:47,770 --> 00:07:52,150 or virtual instances and it will not install any software. 167 00:07:52,150 --> 00:07:56,050 It really will only take care about managing these pods 168 00:07:56,050 --> 00:07:58,540 and the containers in there monitoring them, 169 00:07:58,540 --> 00:08:01,630 it will take care about scaling and about ensuring 170 00:08:01,630 --> 00:08:04,090 that they can be accessed and reached. 171 00:08:04,090 --> 00:08:07,270 It will keep your application up and running. 172 00:08:07,270 --> 00:08:10,970 For creating the resources needed by your application, 173 00:08:10,970 --> 00:08:13,320 you either need to do that on your own, 174 00:08:13,320 --> 00:08:15,760 or you use a tool like Kubermatic 175 00:08:15,760 --> 00:08:18,850 or a managed service like EKS. 176 00:08:18,850 --> 00:08:22,500 And we will see that managed service in action later. 177 00:08:22,500 --> 00:08:23,550 For the moment, 178 00:08:23,550 --> 00:08:25,210 I just want you to be aware of this 179 00:08:25,210 --> 00:08:26,330 and for the moment, 180 00:08:26,330 --> 00:08:29,510 we'll set up a Kubernetes environment locally. 181 00:08:29,510 --> 00:08:33,169 So we'll set up a worker and master node locally 182 00:08:33,169 --> 00:08:34,039 and we'll ensure 183 00:08:34,039 --> 00:08:37,010 that we'll have all the software Kubernetes needs 184 00:08:37,010 --> 00:08:39,663 on these nodes installed locally.