1

00:00:00,450 --> 00:00:06,750

Hi and welcome to this solution video for the practice test of rolling updates and rollbacks



2

00:00:09,390 --> 00:00:13,740

so in this lab we have a total of 13 questions for the first one.



3

00:00:13,740 --> 00:00:20,250

We have deployed a simple web application inspect the parts and services wait for the application to



4

00:00:20,250 --> 00:00:24,810

be fully deployed and you can view the application using the link about the terminal



5

00:00:30,770 --> 00:00:38,120

so if you run a cube CDL get parts we see that four parts have been created for us using the deployment



6

00:00:38,120 --> 00:00:46,100

front end which in turn uses the replica set and as you can see it has four desired and four ready pods



7

00:00:47,300 --> 00:00:55,090

so now let's click on the above portal and here you can see a blue webpage hello from the front.



8

00:00:55,220 --> 00:00:59,150

This is the name of the pod and the application vision as we won



9

00:01:02,910 --> 00:01:11,110

what is the current color of diva application so we know that it's blue run the script named cool hyphen



10

00:01:11,110 --> 00:01:19,590

test to send multiple requests to test diva application take a note of the output so that as a cool



11

00:01:19,590 --> 00:01:29,280

life and test out as each shell script which is in the slash root directory so when we are on that we



12

00:01:29,280 --> 00:01:32,580

can see that the application was and we want color as equal to blue



13

00:01:36,120 --> 00:01:40,010

inspect the deployment and identify the number of parts deployed by it.



14

00:01:40,400 --> 00:01:47,840

We know that it's for what container images used to deploy the applications.



15

00:01:50,910 --> 00:01:58,150

To do this we will run the cube it'll describe command so we are running a describe against the deployment



16

00:01:58,150 --> 00:02:05,150

called front end and here we can see that the image is called Cloud slash whereby font color vision



17

00:02:05,150 --> 00:02:05,450



18

00:02:11,190 --> 00:02:19,110

inspect the deployment and identify the current strategy so if we go to the governor is documentation



19

00:02:19,260 --> 00:02:25,170

for deployment and read up about the strategy you'll see that there are two types of strategy available



20

00:02:25,220 --> 00:02:31,410

want us to recreate or rolling update the default value is falling update



21

00:02:35,100 --> 00:02:40,540

and here within the described command for the deployment you can look for strategy type and it states



22

00:02:40,540 --> 00:02:41,920

here that it's rolling update



23

00:02:46,990 --> 00:02:50,430

if you were to upgrade the application now what would happen.



24

00:02:50,740 --> 00:02:58,870

So if you have control the teary lecture you would know that the all the parts would not be taken down



25

00:02:59,140 --> 00:03:04,990

at the same time when we use a upgrade of the specific deployment.



26

00:03:05,080 --> 00:03:11,380

So in this case because it said to a rolling update parts are upgraded few at a time so that should



27

00:03:11,380 --> 00:03:12,460

be the correct answer



28

00:03:16,310 --> 00:03:16,760

so far.



29

00:03:16,760 --> 00:03:19,150

Question number 8 let us try to do that.



30

00:03:19,610 --> 00:03:27,160

We will try to upgrade the application by changing the image to quote cloud Web life and color vision



31

00:03:27,240 --> 00:03:35,740

to do not delete or recreate the deployment only said the new image so one of the ways to do this is



32

00:03:35,750 --> 00:03:36,960

to use the cube CTO.



33

00:03:36,970 --> 00:03:38,410

Edit deployment command



34

00:03:50,860 --> 00:03:59,400

so I've located the image here and I'm going to change it to read to and as soon as you make that change



35

00:04:02,170 --> 00:04:09,420

would see that a new port has been spun up and the old ones are being deleted.



36

00:04:26,670 --> 00:04:32,730

Now you will see that some of the code commands are hitting the blue web page which is version 1 and



37

00:04:32,730 --> 00:04:34,480

the other ones are going to green.



38

00:04:34,500 --> 00:04:38,790

This is because the strategy is set to rolling update.



39

00:04:38,830 --> 00:04:44,800

Few of them are going to the new pod which has been deployed whereas some of the other requests are



40

00:04:44,800 --> 00:04:50,290

going to the old one and this would be done until all the ports has been replaced by the new image



41

00:04:54,570 --> 00:04:58,200

up to how many ports can be down for the grid at a time.



42

00:04:58,200 --> 00:05:02,160

Consider the current strategy settings and the number of pods is equal to four



43

00:05:05,450 --> 00:05:12,170

so let's run the cube still describe deployment again and you would see that the ruling updates strategy



44

00:05:12,920 --> 00:05:19,940

mentions 25 percent max unavailable which means that 25 percent of the pods can be unavailable at any



45

00:05:19,940 --> 00:05:22,720

given point of time during the upgrade.



46

00:05:22,760 --> 00:05:31,570

So since the maximum number of pods is for the number of pauses that can go down this one



47

00:05:36,670 --> 00:05:40,150

Now let's change the deployment strategy to recreate



48

00:05:46,880 --> 00:05:57,910

let's run the cutesy deal edit deployment again and here will have to change the strategy type to recreate



49

00:06:01,360 --> 00:06:06,270

we can remove the section related to ruling object because that's not relevant for type recreate



50

00:06:09,540 --> 00:06:14,720

and now we'll see if this file.



51

00:06:14,910 --> 00:06:16,530

Now if you're on describe



52

00:06:20,960 --> 00:06:29,090

you'll see that the strategy's diaper set to recreate.



53

00:06:29,180 --> 00:06:33,060

Now let's try to upgrade the deployment with a new image.



54

00:06:33,170 --> 00:06:35,500

Vision we three.



55

00:06:35,650 --> 00:06:37,080

Same as before.



56

00:06:37,420 --> 00:06:39,000

We will use the Edit Come on



57

00:06:48,310 --> 00:06:50,490

we'll change the images to version 3.



58

00:06:56,400 --> 00:07:02,660

Now as soon as we make that change you'll see that all the ports are being terminated of vision V2



59

00:07:07,610 --> 00:07:12,950

which means that this web page will not be available until the new part has been spun up



60

00:07:21,350 --> 00:07:28,120

and if you run the call test you'll see a lot of these requests are getting field nor does the failures



61

00:07:28,130 --> 00:07:33,150

wait for the new application to be really.



62

00:07:33,240 --> 00:07:40,330

Now we have completed the quiz but let's wait for this group to return at least once successful return.



63

00:07:40,420 --> 00:07:41,280

So here we go.



64

00:07:41,290 --> 00:07:50,400

We have got a few pots which are up which means that the call is working and we have got a successful



65

00:07:50,400 --> 00:07:51,020

Web page.



66

00:07:56,720 --> 00:08:04,020

So now all the four pots are now running straight and this is running the wash and we three of the image.



67

00:08:04,200 --> 00:08:05,430

That's it for this video.



68

00:08:05,430 --> 00:08:11,100

Thank you for joining me on this solution lab and I hope you find it useful.



69

00:08:11,100 --> 00:08:12,270

I'll see you in the next one.