WEBVTT

00:00.660 --> 00:02.040
-: One thing that we have not discussed

00:02.040 --> 00:03.120
in the docker world just yet

00:03.120 --> 00:05.880
is how to deal with containers that crash

00:05.880 --> 00:07.320
for some given reason.

00:07.320 --> 00:09.150
So it's entirely possible that you and I

00:09.150 --> 00:10.500
might be running a server

00:10.500 --> 00:12.210
of some sort inside of a container

00:12.210 --> 00:14.760
and maybe that server experiences some error

00:14.760 --> 00:17.250
that cause it to hang or crash.

00:17.250 --> 00:18.750
In this section, we're gonna start taking a look

00:18.750 --> 00:19.710
at a couple of different ways

00:19.710 --> 00:21.570
that we can mitigate that behavior

00:21.570 --> 00:24.540
and figure out how to essentially restart a container

00:24.540 --> 00:28.200
when the software inside of it has an error of some sort.

00:28.200 --> 00:30.630
To get started, we're gonna first add in a little bit

00:30.630 --> 00:33.360
of code to our index.js file to make sure

00:33.360 --> 00:36.720
that our server crashes entirely anytime

00:36.720 --> 00:39.060
someone visits our root route.

00:39.060 --> 00:39.893
So we're gonna add in

00:39.893 --> 00:42.630
kind of a arbitrary forced crash this thing

00:42.630 --> 00:44.070
and then figure out how to deal with it

00:44.070 --> 00:46.110
using docker compose.

00:46.110 --> 00:49.260
So to get started inside of my index.js file,

00:49.260 --> 00:53.070
I'm gonna add in a third require statement here at the top.

00:53.070 --> 00:57.070
It's gonna say const process is require process

00:57.986 --> 01:00.690
and then to make sure that my server crashes

01:00.690 --> 01:03.720
anytime someone tries to access the root route,

01:03.720 --> 01:06.240
I'll find the root route handler right here.

01:06.240 --> 01:08.830
I'll add a new line inside the function body

01:09.810 --> 01:14.430
and I'll say process.exit zero, like so.

01:14.430 --> 01:15.990
Now talk about why we are passing

01:15.990 --> 01:18.000
in the zero right there in just a moment.

01:18.000 --> 01:19.590
But first, let's test this out

01:19.590 --> 01:21.570
and make sure that we can actually get our server

01:21.570 --> 01:25.470
to crash any time that we try to visit this root route.

01:25.470 --> 01:27.660
So I'm gonna flip on over

01:27.660 --> 01:31.770
to my terminal and I'll run docker compose up.

01:31.770 --> 01:33.930
And since we just made a little change

01:33.930 --> 01:36.660
to the code inside one of our images

01:36.660 --> 01:39.540
I'm gonna add on the dash dash build flag

01:39.540 --> 01:42.290
to make sure that we attempt to rebuild that container.

01:43.710 --> 01:46.020
So I'll run that, we're going to rebuild that container

01:46.020 --> 01:47.370
and then everything starts up

01:47.370 --> 01:49.260
and I see my running radar server

01:49.260 --> 01:51.540
and I'm also listening on Port 8081,

01:51.540 --> 01:52.373
but you and I know

01:52.373 --> 01:55.860
that we actually changed the listening port to 4001

01:55.860 --> 01:59.430
inside of our docker composed file right here.

01:59.430 --> 02:02.080
All right, so I'm going to now open up my web browser

02:03.930 --> 02:07.623
and I'm going to navigate to local host colon 4001.

02:08.640 --> 02:09.570
Once I navigate over there,

02:09.570 --> 02:12.540
I'm gonna see an air message like this appear on the screen.

02:12.540 --> 02:14.850
And if I flip back over to my terminal,

02:14.850 --> 02:16.320
you'll notice that it says right here

02:16.320 --> 02:18.660
we exited with code zero.

02:18.660 --> 02:21.510
And so at this point, essentially our running container

02:21.510 --> 02:24.240
and the software inside of it has crashed.

02:24.240 --> 02:26.850
I now want you to try opening up a second tab

02:26.850 --> 02:29.070
and running docker PS.

02:29.070 --> 02:30.090
When we do so you'll notice

02:30.090 --> 02:32.850
that we are now down to just one running container,

02:32.850 --> 02:34.800
just the redis container

02:34.800 --> 02:38.880
and our node server is no longer running whatsoever.

02:38.880 --> 02:41.670
So clearly that thing has crashed

02:41.670 --> 02:43.830
and we might want to automatically restart it

02:43.830 --> 02:45.330
at some point in time.

02:45.330 --> 02:47.040
So let's take a quick pause right here

02:47.040 --> 02:49.590
and take a look at what our different options are

02:49.590 --> 02:50.640
in the next section.

02:50.640 --> 02:53.090
So quick break and I'll see you in just a minute.