1
00:00:01,300 --> 00:00:03,390
I'm taking a look at the listener test file.

2
00:00:03,390 --> 00:00:08,140
Back inside of our Nats test project you'll recall that inside of here we said that anytime that our

3
00:00:08,140 --> 00:00:13,330
process was about to be stopped either by an interrupt signal or a terminate signal we tried to close

4
00:00:13,330 --> 00:00:16,940
down our client gracefully so we would call stand out close.

5
00:00:16,990 --> 00:00:18,750
The client would then attempt to close.

6
00:00:18,790 --> 00:00:21,590
Once it did close it would he met the closed event.

7
00:00:21,610 --> 00:00:23,610
So we captured that closed event.

8
00:00:23,620 --> 00:00:26,170
We said OK the thing has been cleaned up successfully.

9
00:00:26,170 --> 00:00:29,080
And then we actually exited our process.

10
00:00:29,080 --> 00:00:34,780
So we're gonna do something very similar back inside of our ticketing project as well so let me first

11
00:00:34,780 --> 00:00:37,060
tell you one way that we could implement this.

12
00:00:37,060 --> 00:00:39,690
This is not the good way to do it it's not the right way.

13
00:00:39,700 --> 00:00:43,660
I just want to point out some issues and I'm going to show it to you because this might be something

14
00:00:43,690 --> 00:00:45,200
you would do on your own.

15
00:00:45,440 --> 00:00:47,480
I'm going to take stand out on closed right here.

16
00:00:47,500 --> 00:00:48,780
I'm going to copy it.

17
00:00:48,910 --> 00:00:51,340
Go back over to Nats wrapper.

18
00:00:51,520 --> 00:00:57,130
I'm gonna find the Connect function and then right after that I'm going to paste in that stand out on

19
00:00:57,130 --> 00:00:57,890
close.

20
00:00:57,940 --> 00:01:02,350
I do have to change stand to this dot underscore a client of course.

21
00:01:02,440 --> 00:01:03,820
I'm also gonna go back to listener.

22
00:01:03,820 --> 00:01:09,740
I'm gonna take the to process dot on statements at the bottom copy those look back over it and that's

23
00:01:09,740 --> 00:01:15,630
wrapper and paste them right underneath and then once again I'll update stand to this dot underscore

24
00:01:15,720 --> 00:01:21,060
client and I go to put a exclamation or ultimately you know we could just do this dot client be a better

25
00:01:21,060 --> 00:01:21,710
way of doing this.

26
00:01:23,420 --> 00:01:26,550
OK so this right here would definitely work.

27
00:01:26,840 --> 00:01:32,660
This would make sure that after we connect we then attempt to capture any closed event.

28
00:01:32,660 --> 00:01:37,310
And once we capture it we're going to make sure we do cuts logs say hey everything was shut down successfully

29
00:01:37,310 --> 00:01:38,430
and then exit.

30
00:01:38,450 --> 00:01:43,380
And we also set up two listeners on the process to make sure there's ever an interrupt or terminate.

31
00:01:43,470 --> 00:01:46,100
We are going to go ahead and close everything down.

32
00:01:46,130 --> 00:01:48,230
This would definitely work without a doubt.

33
00:01:48,350 --> 00:01:50,210
But we do not want to do this.

34
00:01:50,210 --> 00:01:55,880
The reason why is that we have some method inside of here that is buried inside of some class inside

35
00:01:55,880 --> 00:02:02,080
of our project that can possibly cause our entire process to exit entirely.

36
00:02:02,110 --> 00:02:07,390
We do not ever want to have some method inside of some hidden away file that is going to suddenly cause

37
00:02:07,390 --> 00:02:12,940
the entire program to exit definitely bad design right here as we might write this file.

38
00:02:12,990 --> 00:02:18,150
We might decide to say put it inside of our common module and then use it throughout all of our different

39
00:02:18,150 --> 00:02:19,490
services.

40
00:02:19,680 --> 00:02:23,720
We would then have some shared library that might cause our program to exit entirely.

41
00:02:23,730 --> 00:02:29,300
Can you imagine what that would be like if Mongoose behave the same if Mongoose could just suddenly

42
00:02:29,300 --> 00:02:32,120
arbitrarily decide to exit your entire program.

43
00:02:32,120 --> 00:02:34,640
That would probably not be super ideal.

44
00:02:34,670 --> 00:02:39,320
So rather than putting code like this hidden away inside of some method we only ever want to allow something

45
00:02:39,320 --> 00:02:46,830
to exit our entire program from inside of some more central location such as our index t s file so rather

46
00:02:46,830 --> 00:02:51,870
than trying to directly exits from inside of connect we're going to instead put together a little bit

47
00:02:51,870 --> 00:02:56,580
of functionality that's going to allow us to see that the client is attempting to close and then we'll

48
00:02:56,580 --> 00:03:00,740
do the actual exit command from inside of index start to yes.

49
00:03:00,860 --> 00:03:03,920
So again this would work but not ideal.

50
00:03:03,940 --> 00:03:05,170
So how are we going to fix this.

51
00:03:05,170 --> 00:03:08,320
Well essentially we're just going to take the same code where we're going to write it out inside of

52
00:03:08,320 --> 00:03:09,890
our index not test file.

53
00:03:10,060 --> 00:03:11,710
I'm going to cut all that right there.

54
00:03:13,870 --> 00:03:20,140
I'm then going to go back to index start to yes and then right after our connect statement I'm going

55
00:03:20,140 --> 00:03:25,360
to paste all that stuff in so we do have to repair a couple of references.

56
00:03:25,430 --> 00:03:33,860
I'm to make sure I update this to Nat's wrapper client and then I'll update this right here.

57
00:03:33,890 --> 00:03:41,460
Both these this clients both them should now be at Nats wrapper not clients like so OK.

58
00:03:41,490 --> 00:03:42,590
So this is much better.

59
00:03:42,600 --> 00:03:47,040
So now instead of a very central location inside of our project we can see that there is a condition

60
00:03:47,040 --> 00:03:50,320
under which we're going to exit our entire program.

61
00:03:50,400 --> 00:03:55,620
It would be really nice if we could somehow integrate this into our class because right now.

62
00:03:55,620 --> 00:04:00,780
Well you know this is some very repetitive logic right here that we're probably going to want to duplicate

63
00:04:00,780 --> 00:04:02,240
between our different services.

64
00:04:02,310 --> 00:04:07,760
But again it just would not be a great form to hide this way inside of some far away file.

65
00:04:07,830 --> 00:04:12,300
So this is now the big question here is how do we actually test this stuff and make sure it's working

66
00:04:12,330 --> 00:04:13,500
as expected.

67
00:04:13,500 --> 00:04:15,870
Well let me show you how we can test this very easily.

68
00:04:15,870 --> 00:04:20,790
The first thing we're going to do is go back over to our terminal and we are going to re start scaffold.

69
00:04:20,790 --> 00:04:24,720
Now it's really important that you restart scaffold because if you don't you're going to run into some

70
00:04:24,780 --> 00:04:26,820
really strange situation.

71
00:04:26,880 --> 00:04:29,460
So I'm going to kill scaffold.

72
00:04:29,620 --> 00:04:31,070
I'm gonna see everything deleted.

73
00:04:31,070 --> 00:04:33,110
I'm going to give those pods just a moment.

74
00:04:34,170 --> 00:04:35,060
To get cleaned up.

75
00:04:35,200 --> 00:04:36,930
So looks like they're all getting terminated.

76
00:04:37,330 --> 00:04:37,610
Yep.

77
00:04:37,640 --> 00:04:38,690
Good enough.

78
00:04:38,750 --> 00:04:40,430
I'll go ahead and then restart scaffold

79
00:04:43,960 --> 00:04:46,490
so it's going to take a second for everything to come back online

80
00:04:49,100 --> 00:04:51,180
while that is all booting up.

81
00:04:51,590 --> 00:04:53,320
There's my connected to that's right there by the way.

82
00:04:53,330 --> 00:04:57,980
There's my connect connected to mongo DV and listening on port three thousand cats everything is now

83
00:04:58,010 --> 00:04:59,770
refreshed all running.

84
00:04:59,810 --> 00:05:03,300
Let's now try to simulate scarf or sweep.

85
00:05:03,300 --> 00:05:08,410
We're gonna try to simulate gnats going down we're gonna try to just delete the Nats pod entirely.

86
00:05:08,410 --> 00:05:12,580
That is going to cause the client to emit its own close event.

87
00:05:12,580 --> 00:05:18,750
So when that thing gets closed we should then see the process exit entirely so I going to go back to

88
00:05:18,780 --> 00:05:19,420
my terminal.

89
00:05:19,470 --> 00:05:21,690
I to find a new terminal window.

90
00:05:21,860 --> 00:05:23,990
I'm gonna get a listing of all my pods.

91
00:05:23,990 --> 00:05:29,780
I'm gonna find a pod running that's and then do a cube TTL delete pod and then put in the name of that

92
00:05:29,780 --> 00:05:33,050
pod if I now go backwards.

93
00:05:33,050 --> 00:05:38,000
The primary scaffold window I will see eventually a console log of Nats connection closed.

94
00:05:38,000 --> 00:05:39,160
Perfect.

95
00:05:39,230 --> 00:05:44,150
And then a couple seconds after that for you it might take several seconds might take all that to 30

96
00:05:44,150 --> 00:05:45,350
seconds or a minute.

97
00:05:45,350 --> 00:05:52,740
You should eventually see the container started back up so we did successfully kill the process inside

98
00:05:52,740 --> 00:05:58,310
there by calling process exit so we exited the process behind the scenes.

99
00:05:58,340 --> 00:06:04,340
The pod running our container saw that the primary process of our container exited the pod then automatically

100
00:06:04,340 --> 00:06:07,410
decided to restart the container inside there.

101
00:06:07,460 --> 00:06:11,460
So it restarted the container and that's why we are seeing brand new output right here.

102
00:06:12,260 --> 00:06:17,380
From T.S. no def we can probably also confirm that the container inside that pod was restarted.

103
00:06:17,410 --> 00:06:24,670
If we do another K or cube Seitel get pods and now we'll see that our ticket's deployment right there

104
00:06:24,670 --> 00:06:29,990
has had one restart that restart is because we lost our connection to net's all right.

105
00:06:29,990 --> 00:06:32,470
So I'd say this is looking pretty good.

106
00:06:32,510 --> 00:06:36,920
We are going to exit our process entirely any time that we lose our connection to gnats.

107
00:06:36,920 --> 00:06:39,130
That is definitely something we want.

108
00:06:39,170 --> 00:06:43,350
We are also going to make sure that any time that we get this interrupt signal or terminate signal we're

109
00:06:43,350 --> 00:06:45,940
going to close down the client manually as well.

110
00:06:45,980 --> 00:06:46,960
This looks pretty great.

111
00:06:46,970 --> 00:06:48,080
Let's take a pause right here.

112
00:06:48,080 --> 00:06:52,250
We still also have to test and make sure that whenever we create a new ticket we published an event

113
00:06:52,250 --> 00:06:52,690
as well.

114
00:06:52,820 --> 00:06:54,710
So still a little bit of stuff for us to do.