1
00:00:00,001 --> 00:00:01,767
[No Audio]

2
00:00:01,767 --> 00:00:04,734
In this demo, we will perform a few more

3
00:00:04,734 --> 00:00:06,767
orchestration related tasks.

4
00:00:08,167 --> 00:00:12,867
If you remember clearly, our service web-server had 3 replicas

5
00:00:12,867 --> 00:00:18,567
of latest: nginx image. Let's scale our service and

6
00:00:18,567 --> 00:00:22,734
increase its number of replicas to 6. We can do

7
00:00:22,734 --> 00:00:26,334
it with docker service scale command, followed by

8
00:00:26,334 --> 00:00:30,267
the name of service, and the number of replicas.

9
00:00:30,734 --> 00:00:34,034
Once Swarm has verified the scaling, we can do it

10
00:00:34,034 --> 00:00:37,234
as well using docker service ps, followed by the

11
00:00:37,234 --> 00:00:41,201
service name. And as you can see, instead of 3,

12
00:00:41,201 --> 00:00:44,234
now we have 6 containers running on nginx

13
00:00:44,234 --> 00:00:47,067
latest image, 3 of them are scheduled on

14
00:00:47,067 --> 00:00:50,201
manager and 3 of them are scheduled on worker.

15
00:00:50,334 --> 00:00:53,234
All of those 6 are in running state, and 3

16
00:00:53,234 --> 00:00:55,601
of them seem to be quite new.

17
00:00:56,267 --> 00:00:58,134
As you might have expected,

18
00:00:58,134 --> 00:01:00,967
if we run docker ps -a, on both

19
00:01:00,967 --> 00:01:04,434
manager and worker-1, we will see 3

20
00:01:04,434 --> 00:01:06,667
containers running on each of them.

21
00:01:07,401 --> 00:01:12,934
While worker-1 has two new containers, manager has 1 new container.

22
00:01:12,934 --> 00:01:15,967
Furthermore, we can even roll out some

23
00:01:15,967 --> 00:01:19,001
updates on all of these 6 containers.

24
00:01:19,501 --> 00:01:24,734
As you know, all of these containers are running on nginx:latest image,

25
00:01:24,734 --> 00:01:27,801
we can change it to nginx:alpine.

26
00:01:27,801 --> 00:01:30,867
If you're wondering what is the difference? Well, the

27
00:01:30,867 --> 00:01:34,434
latest version of nginx is built on top of

28
00:01:34,434 --> 00:01:38,867
Ubuntu base image, whereas alpine version is based on

29
00:01:38,867 --> 00:01:42,567
top of minimal Alpine Linux image. Let's use the

30
00:01:42,567 --> 00:01:45,567
command docker service update followed by what

31
00:01:45,567 --> 00:01:48,667
kind of field do we want to update. We want to

32
00:01:48,701 --> 00:01:52,001
update the image of the service. Once we hit

33
00:01:52,001 --> 00:01:56,467
enter, all of these tasks get updated one at a time.

34
00:01:56,467 --> 00:01:59,434
Once the update process is complete, we can

35
00:01:59,434 --> 00:02:02,667
verify it with docker service inspect command

36
00:02:02,667 --> 00:02:05,634
and let's make sure that the result of the inspect

37
00:02:05,634 --> 00:02:09,567
command is pretty printed. As you can see, the

38
00:02:09,567 --> 00:02:12,167
service mode is still replicated, the number of

39
00:02:12,167 --> 00:02:15,367
replicas is 6. If we go to the container

40
00:02:15,367 --> 00:02:19,534
specification, instead of showing nginx:latest, it

41
00:02:19,534 --> 00:02:22,767
shows nginx:alpine, which means all of the

42
00:02:22,767 --> 00:02:25,934
containers are switched from nginx:latest to

43
00:02:25,934 --> 00:02:29,567
alpine image. Finally, we can remove our service

44
00:02:29,567 --> 00:02:32,601
using docker service rm command followed by the

45
00:02:32,601 --> 00:02:37,467
name of the service. And as a notification, we get

46
00:02:37,467 --> 00:02:41,834
the name of service. Let's type docker ps -a,

47
00:02:41,834 --> 00:02:45,434
and as you can see, every container is being

48
00:02:45,434 --> 00:02:50,134
taken down one by one. If we do the same on worker-1 node,

49
00:02:50,134 --> 00:02:52,734
you will see that all of the containers

50
00:02:52,734 --> 00:02:56,467
are taken down and also removed. Let's wait for a

51
00:02:56,467 --> 00:03:00,001
while and use the same command on manager as well.

52
00:03:00,434 --> 00:03:03,934
Well, now even our manager is empty.

53
00:03:04,234 --> 00:03:08,967
Finally, let's clean up our cluster by making sure that

54
00:03:08,967 --> 00:03:13,067
worker node is also leaving the cluster. Just like

55
00:03:13,067 --> 00:03:17,101
we did with worker-2, we will also make worker-1

56
00:03:17,101 --> 00:03:21,067
leave the cluster voluntarily using docker swarm leave command.

57
00:03:21,067 --> 00:03:25,701
We are back to having one Docker Host which is manager.