1 00:00:00,490 --> 00:00:04,510 This is the solution video for the classroom maintenance lab. 2 00:00:05,720 --> 00:00:09,950 So for the first question, let us explore the environment first, how many news you see in the classroom? 3 00:00:09,960 --> 00:00:12,470 So let's run cubicle, get notes. 4 00:00:16,770 --> 00:00:19,290 And we can see that this is a whole new cluster. 5 00:00:26,530 --> 00:00:29,330 How many applications do you see hosted on the cluster? 6 00:00:29,530 --> 00:00:31,040 Check the number of deployments. 7 00:00:31,060 --> 00:00:33,940 So let's try and Cube still get deployment. 8 00:00:39,320 --> 00:00:45,380 Kubek, you get deployment's and we can see that there are two deployments by the name of blue and red, 9 00:00:45,600 --> 00:00:46,880 the default namespace. 10 00:00:49,000 --> 00:00:51,190 On which news are the applications holstered? 11 00:00:54,620 --> 00:01:01,350 So let's run the command to get parts dash or wait to see the north and with the running. 12 00:01:01,380 --> 00:01:04,790 So in this case, we can see that they are running on another one and no route to. 13 00:01:08,300 --> 00:01:13,640 You need to take notes, rule one out for maintenance and note of all applications and it on schedule. 14 00:01:14,360 --> 00:01:17,990 So for this will run CubeSat Al Green. 15 00:01:23,850 --> 00:01:25,590 Node 01. 16 00:01:28,040 --> 00:01:33,420 So they should drain all the ports which are running on the spot, but will have to use dash, dash, 17 00:01:33,470 --> 00:01:38,100 ignore them, it's because there are some ports which are part of a demand set here. 18 00:01:38,870 --> 00:01:41,640 So without that, it will not drain the node. 19 00:01:48,900 --> 00:01:51,380 So you can see that the parts are getting evicted. 20 00:01:51,480 --> 00:01:54,090 That means that the train is working well. 21 00:01:55,740 --> 00:02:02,520 So soon it should not be hosting any parts other than the new ones which we asked to come on to ignore. 22 00:02:03,870 --> 00:02:05,460 What news are the apps on now? 23 00:02:05,470 --> 00:02:11,760 So let's on the same comment as before, which is cubicle get Pott's Dash all white, and now we can 24 00:02:11,760 --> 00:02:14,880 see that they are on NEWSROOM two and NEWSROOM three. 25 00:02:22,120 --> 00:02:28,660 So now let's wait for the command of the question to load up the maintenance task have been completed, 26 00:02:28,660 --> 00:02:30,700 configured in order to reschedule again. 27 00:02:31,790 --> 00:02:40,100 So let's run the command cube, dear, uncowed on, not zero one, so this will make sure that it is 28 00:02:40,100 --> 00:02:45,940 scheduled again and we can verify that Peranich still get notes so we can see that no one is ready now. 29 00:02:46,900 --> 00:02:51,500 So it is ready to schedule new ports, how many ports are scheduled, another one. 30 00:02:51,520 --> 00:02:52,240 So let's run. 31 00:02:52,810 --> 00:02:57,310 It will get ports Stachel wide again and see if there are any ports which are running on that one and 32 00:02:57,310 --> 00:02:59,530 looks like there aren't so. 33 00:03:00,660 --> 00:03:06,510 There are no new ports, no ports which are running on Luzira, one, why are there no ports are not 34 00:03:06,510 --> 00:03:06,870 the one. 35 00:03:06,870 --> 00:03:12,840 That's because we ran the train command, which affected the ports and rescheduled them on a different 36 00:03:12,840 --> 00:03:13,260 note. 37 00:03:14,290 --> 00:03:19,450 And as a result, until a new board has been scheduled, it will not run on one. 38 00:03:21,980 --> 00:03:24,640 Why are there no pause plays on the emotional. 39 00:03:25,460 --> 00:03:32,060 So let's take these teams, which are set on the snow so far that we could still describe a. monster 40 00:03:32,480 --> 00:03:33,870 and grab 14. 41 00:03:34,460 --> 00:03:40,460 So by default, Martineau's would have teed off no schedule, which will prevent any new ports to reschedule 42 00:03:40,460 --> 00:03:41,620 on the monster itself. 43 00:03:42,760 --> 00:03:48,100 It is now time to take down the road to for maintenance before you remove all workload's from loser, 44 00:03:48,100 --> 00:03:52,620 to answer the following question, can you drain on the road to using the same command as NEWSROOM? 45 00:03:52,770 --> 00:03:53,680 But let's try that. 46 00:03:54,630 --> 00:03:57,630 So let's try and Cube Carlgren, not zero to. 47 00:03:58,820 --> 00:04:01,930 And let's also make use of the dash dash, ignore different sets. 48 00:04:01,970 --> 00:04:05,090 And in this case, it's failing because of an error. 49 00:04:05,150 --> 00:04:12,560 It seems that there are parts on NEWSROOM two which are not managed by demons, head replication controller 50 00:04:12,560 --> 00:04:18,970 or Percocet or a state would have to use dash dash forced to override and the name of the partners HRR 51 00:04:18,980 --> 00:04:19,760 Dash app. 52 00:04:22,440 --> 00:04:24,350 So the answer is no, you must Forcett. 53 00:04:25,170 --> 00:04:26,360 Why do you need to force it? 54 00:04:26,430 --> 00:04:31,170 We already know that there's a part which is not part of the replica said. 55 00:04:32,140 --> 00:04:35,980 What is the name of the boat we already saw that as a child app. 56 00:04:39,680 --> 00:04:46,400 What would happen to a app if not through to a screen, forcefully so, because it's not part of a replica 57 00:04:46,400 --> 00:04:52,130 set or a communication controller or demon set on stateful set, it will be lost over. 58 00:04:54,630 --> 00:04:56,890 Reynolds Rotu and mark it on schedule. 59 00:04:57,600 --> 00:04:59,550 So let's use Dashty force now. 60 00:05:01,210 --> 00:05:07,270 And here you can see that the part, a app has been evicted as well, and it will not be rescheduled 61 00:05:07,270 --> 00:05:09,880 on any other note because it's not managed by. 62 00:05:10,930 --> 00:05:13,810 And you have the replica sets, all replication controllers. 63 00:05:16,400 --> 00:05:22,030 He has our critical applications, we do not want to schedule any more absolute zero three zero three 64 00:05:22,030 --> 00:05:25,720 as unscheduled, but do not remotely currently running on it. 65 00:05:25,720 --> 00:05:29,200 So we do not want to drain modulatory in this case. 66 00:05:29,890 --> 00:05:34,930 What we want to do is called on it instead so that it does not accept any more points to reschedule 67 00:05:34,930 --> 00:05:35,200 on a. 68 00:05:37,970 --> 00:05:39,500 So if Iran could still get. 69 00:05:40,960 --> 00:05:46,870 Part, I get notes, we'll see that newsroom tree is ready, but she really is disabled. 70 00:05:48,230 --> 00:05:54,660 Nusreta was drained earlier, so it was a liquid state and that was the last question of this lab. 71 00:05:54,700 --> 00:05:56,310 Thank you for joining me in this video. 72 00:05:56,370 --> 00:05:57,590 I'll see you in the next one.