1 00:00:00,150 --> 00:00:08,670 In this video, I will demonstrate upgrading a commonalties cluster deployed using Badme, so for this 2 00:00:08,670 --> 00:00:15,780 as a prerequisite, you need a cabinet is deployed using the cube admin tool and it could be a local 3 00:00:15,780 --> 00:00:22,710 set up on your laptop or one on the cloud, or you could be using our labs in this demo. 4 00:00:22,720 --> 00:00:28,070 I happen to be using a cluster from the playground here. 5 00:00:28,500 --> 00:00:34,550 So if you just go to Kalakuta Answers for Covenant's playground, you can instantly open up a lab session, 6 00:00:34,920 --> 00:00:41,990 a playground which has a master and a worker node, and it's just very easy to practice cluster upgrades. 7 00:00:42,600 --> 00:00:49,260 So on the right here, I have a cabinet cluster, um, terminal to accompany this cluster on Kalakuta 8 00:00:49,260 --> 00:00:50,760 with a mass ranneberger node. 9 00:00:50,970 --> 00:00:57,150 The master node, which is the control plane is at the top and the worker node is the terminal to the 10 00:00:57,150 --> 00:00:57,360 worker. 11 00:00:57,360 --> 00:01:02,600 Notice below and on the left here I have the component is documentation pages open. 12 00:01:03,450 --> 00:01:09,030 Now, if you are not sure how to find this page, just search for Cube Ethereum upgrade in the search 13 00:01:09,030 --> 00:01:09,270 bar. 14 00:01:09,270 --> 00:01:12,510 And the first link is what you are looking for. 15 00:01:13,200 --> 00:01:19,200 And by default, it opens the page with the latest instructions on upgrading, which as of now is from 16 00:01:19,200 --> 00:01:22,020 one nine one point nineteen to one to 20. 17 00:01:22,500 --> 00:01:24,630 And my cluster is that one of the 18. 18 00:01:24,630 --> 00:01:31,950 So I will choose the instructions for upgrading from one of 18 to 19 by clicking this link. 19 00:01:32,790 --> 00:01:36,620 So choose the relevant documentation page depending on your requirements. 20 00:01:37,170 --> 00:01:43,440 Now, once the environment is loaded, I run the launch launch site script to wait for the cluster to 21 00:01:43,440 --> 00:01:43,830 be ready. 22 00:01:43,830 --> 00:01:50,220 But remember that this is only applicable to a Kitakata, so you might not need this in your if you're 23 00:01:50,220 --> 00:01:51,870 using it in another environment. 24 00:01:52,020 --> 00:01:57,960 So this particular step is only relevant to, uh, environments provisioned with Kalakuta. 25 00:01:58,620 --> 00:02:00,300 And so it just waits. 26 00:02:00,510 --> 00:02:03,860 It just creates like just a cluster and wait for it to come up. 27 00:02:04,230 --> 00:02:11,700 So you make sure the cluster is ready and by running the queue to get nodes command, it shows the control 28 00:02:11,700 --> 00:02:13,440 plane is not ready. 29 00:02:13,440 --> 00:02:15,090 So let's just wait for it to be ready. 30 00:02:15,660 --> 00:02:19,050 And it doesn't even show the node right now. 31 00:02:19,050 --> 00:02:24,780 So we'll just wait for the the control plane and the node to be in real estate so we could use the watch 32 00:02:24,780 --> 00:02:30,570 command with the cubicle, get node command to wait for both nodes to be ready. 33 00:02:31,530 --> 00:02:35,610 So in this setup, I have a single control plane and a single worker node. 34 00:02:42,350 --> 00:02:44,270 OK, so now what the notes are ready. 35 00:02:46,160 --> 00:02:52,520 And you could run any cube Adam command to verify that curium is installed and that the cluster was 36 00:02:52,520 --> 00:02:54,700 set up using BDM. 37 00:02:56,120 --> 00:03:05,540 So I run the token list command and it shows a token, so that kind of shows that it used that token 38 00:03:05,540 --> 00:03:11,870 to set up the cluster and you could also run the cubicle, get pods command with the Daschle option 39 00:03:12,380 --> 00:03:15,770 to list all the parts, including those in the cube system namespace. 40 00:03:16,100 --> 00:03:22,730 And we see that there is the core DNS, the Exelis control plane, and the networking solution is flannel. 41 00:03:23,290 --> 00:03:29,240 In this demo, we're going to stick to just upgrading the capabilities components only and we will not 42 00:03:29,240 --> 00:03:29,780 upgrade it. 43 00:03:29,780 --> 00:03:31,510 Flannel networking solution. 44 00:03:31,520 --> 00:03:34,550 So there are separate instructions for upgrading those. 45 00:03:34,760 --> 00:03:39,290 We will just stick to upgrading the control plane as well as the worker notes. 46 00:03:40,450 --> 00:03:45,250 So run the cubicle, get notes, command to verify the current version of the cluster, so it shows 47 00:03:45,250 --> 00:03:47,500 as version one, dot 18, dot zero. 48 00:03:48,370 --> 00:03:51,700 But remember that this is a version of the cube that running on the note. 49 00:03:52,660 --> 00:03:59,500 The cube admin, too, will help us see the exact version of all the different components when we run 50 00:03:59,500 --> 00:04:00,960 the the plan command. 51 00:04:01,010 --> 00:04:07,930 So in the documentation page here on the left, the first step is to determine which version that is 52 00:04:07,930 --> 00:04:09,180 it that you want to upgrade to. 53 00:04:09,760 --> 00:04:15,200 We know that is one point nineteen, but we don't know exactly the exact version within one point nineteen. 54 00:04:15,790 --> 00:04:17,890 So for this, we must run these commands. 55 00:04:17,890 --> 00:04:25,000 And before you do that, if you are not sure which operating system you're on, you could run the cat 56 00:04:25,810 --> 00:04:32,710 ATC slash star, release our command and it list the contents of the release file. 57 00:04:32,710 --> 00:04:35,880 And you will see that in this case it is Ubuntu. 58 00:04:36,670 --> 00:04:45,310 So we'll stick to the Ubuntu commands throughout this throughout this demo, we run the app commands 59 00:04:45,310 --> 00:04:48,060 listed here to see the different versions available for upgrade. 60 00:04:48,070 --> 00:04:53,860 So since we are upgrading to one point nineteen, we will choose one that 19 six version, which happens 61 00:04:53,860 --> 00:04:56,440 to be the latest in one point nineteen. 62 00:04:56,980 --> 00:04:58,390 So just note that down. 63 00:04:59,490 --> 00:05:06,390 So the high level process is to upgrade the control plane load and then the upgrade CubeSat and you 64 00:05:06,400 --> 00:05:10,020 cuddle on those notes and then upgrade worker notes. 65 00:05:11,190 --> 00:05:17,640 So we will follow this procedure, but if you only want to upgrade certain parts, such as the work 66 00:05:17,640 --> 00:05:23,430 of along, then you can skip this initial section where which is used for upgrading the control plane. 67 00:05:23,940 --> 00:05:28,230 Or if you just want to upgrade the control plane and skip the working out section, you could do that 68 00:05:28,230 --> 00:05:28,560 as well. 69 00:05:28,590 --> 00:05:36,080 So just so feel free to choose the section that you want that you really want to do. 70 00:05:36,960 --> 00:05:42,630 So let's first proceed with upgrading the control plane, so the first step is to upgrade Zubayda itself 71 00:05:42,630 --> 00:05:43,620 on the control plane. 72 00:05:43,650 --> 00:05:49,890 And for this, we run this set of commands to update and install new versions of QUIDAM for our target 73 00:05:49,890 --> 00:05:52,980 version, which happens to be one, dot 19, dot six. 74 00:05:53,400 --> 00:05:59,760 So I will copy to a notepad and I just change the version in the command and will execute the command 75 00:05:59,760 --> 00:06:00,480 in the terminal. 76 00:06:01,480 --> 00:06:03,190 We'll wait for that update to finish. 77 00:06:12,000 --> 00:06:17,420 So the update is finished and now we verify the version of Ubaidi, I'm using the Canadian version command 78 00:06:17,970 --> 00:06:21,510 and it seems to be upgraded successfully to one point nineteen dot six. 79 00:06:22,610 --> 00:06:28,170 Next, we look at our upgrade plan, the man running the cube adiam upgrade plan command. 80 00:06:28,660 --> 00:06:33,200 It lists the current version of components which are at one point eighteen dollars zero and the available 81 00:06:33,200 --> 00:06:37,760 versions to upgrade to which is one point nineteen dollars six as desired. 82 00:06:38,730 --> 00:06:44,370 The next step is to run the Cuba upgrade, appli command to upgrade to version one, not six. 83 00:06:45,330 --> 00:06:49,700 So for this, we run the command and then confirm on the prompt. 84 00:06:50,500 --> 00:06:55,650 Now this might take a few minutes, so let's just wait for it to complete, because this is when it's 85 00:06:55,650 --> 00:06:57,210 actually performing the upgrade. 86 00:07:15,510 --> 00:07:20,970 Once completed, we see the message that the cluster was upgraded to one that 19 dots successfully to 87 00:07:20,970 --> 00:07:23,590 verify if you ran the cube. 88 00:07:23,970 --> 00:07:30,300 If you run the cube idiom, upgrade plan, command a game that we see that there are no more upgrade 89 00:07:30,300 --> 00:07:31,230 recommendations. 90 00:07:31,710 --> 00:07:34,950 So that confirms that all the components have been upgraded successfully. 91 00:07:45,760 --> 00:07:51,990 If you run the cubicle, get notes, command, now we will see the version to be one that it is not 92 00:07:52,030 --> 00:07:52,310 zero. 93 00:07:52,550 --> 00:07:59,590 Still, this is because what you see here is a version of Hublot and Aquarium Upgrade Command does not 94 00:07:59,590 --> 00:08:00,820 upgrade to blit. 95 00:08:01,490 --> 00:08:02,780 So that's the next step. 96 00:08:03,100 --> 00:08:10,510 Now, since Kuebler is responsible for running pods, we must drain the node before upgrading CubeSat. 97 00:08:12,590 --> 00:08:19,970 So we run the cubicle drain command to drain the control node, and once done, we proceed with the 98 00:08:19,970 --> 00:08:25,250 upgrade commands to upgrade Kubelik and the cubicle utility itself. 99 00:08:30,610 --> 00:08:37,960 So as before, I copy to a text, file the commands and we change the versions to one point ninety six 100 00:08:37,960 --> 00:08:40,450 and then we execute the commands. 101 00:08:48,470 --> 00:08:52,820 Once done, reload the Neeman and restart Hublot. 102 00:08:54,950 --> 00:09:01,080 When we run the cubicle, get notes command, we see the control plane is now upgraded to understanding 103 00:09:01,130 --> 00:09:04,910 that the status is schedule disabled. 104 00:09:04,940 --> 00:09:09,050 Now, this is because we drained the node earlier, so we must now uncoordinated. 105 00:09:09,050 --> 00:09:11,540 And that's what we will do now for this. 106 00:09:11,540 --> 00:09:13,670 We run the cubicle, unquote, in command. 107 00:09:14,590 --> 00:09:18,860 We verified the status again, and we see that the note is now ready. 108 00:09:19,540 --> 00:09:22,040 So we have successfully upgraded the master node. 109 00:09:22,870 --> 00:09:25,260 Now, the next step is to upgrade the worker notes. 110 00:09:25,840 --> 00:09:31,130 Now, as before, the first step is to upgrade Zubayda M on the worker notes. 111 00:09:31,360 --> 00:09:38,380 So we run the app, install commands to upgrade Kubi a.m. to the target version and note that this is 112 00:09:38,380 --> 00:09:41,670 to be done on the worker node and that is to be updated. 113 00:09:42,160 --> 00:09:44,650 So most of course, most of the commands. 114 00:09:45,760 --> 00:09:51,390 To upgrade worker nodes are done on the worker nodes except the cube related command. 115 00:09:51,610 --> 00:09:57,040 I'll I'll point that out when we reach to that point, but make sure that you are on the appropriate 116 00:09:57,040 --> 00:10:01,480 node and when you're running the respective commands. 117 00:10:04,580 --> 00:10:08,600 So now that is done, the next step is to upgrade Kubla configuration. 118 00:10:10,150 --> 00:10:17,400 We run the command to upgrade that configuration, and once done, the next step is to drain the node 119 00:10:18,070 --> 00:10:20,870 and this is where a lot of people make mistake. 120 00:10:20,890 --> 00:10:25,810 The drain command is to be run on the control node because it's a cubicle command. 121 00:10:26,260 --> 00:10:30,550 If you run it on a worker node, you will get an error like this. 122 00:10:33,090 --> 00:10:39,600 So remember to run it on the control plane, not most of the commands, as I just said, for upgrade, 123 00:10:40,150 --> 00:10:47,150 upgrading or working out a run on the work node, except for the shuttle commands to train and coordinate. 124 00:10:47,580 --> 00:10:50,280 And those are done to be done on the controlled plane. 125 00:10:50,280 --> 00:10:50,670 Not. 126 00:10:56,700 --> 00:10:58,950 So once done, the next step is to upgrade. 127 00:10:59,460 --> 00:11:05,110 And for this, we run the install commands with the target versions of these packages. 128 00:11:05,490 --> 00:11:12,150 And once that is completed, we reload Biman and restart, uh, running the cubicle, get node's. 129 00:11:12,150 --> 00:11:18,740 Command confirms the upgrade is successful on the worker node and scheduling is disabled as we haven't 130 00:11:19,050 --> 00:11:20,100 and couldn't yet. 131 00:11:20,800 --> 00:11:21,960 Uh, so let's do that. 132 00:11:22,590 --> 00:11:28,160 And remember, this has to be done on the master node as well, and that's it. 133 00:11:28,710 --> 00:11:32,630 We have now successfully completed the upgrade procedure. 134 00:11:33,810 --> 00:11:36,530 Well, head over to the labs and try this for yourself.