WEBVTT

0:00:02.660000 --> 0:00:10.260000
 In this video, I'm going to demonstrate
 how to deploy a website to an

0:00:10.260000 --> 0:00:11.700000
 Azure Kubernetes service.

0:00:11.700000 --> 0:00:12.940000
 This is pretty straightforward.

0:00:12.940000 --> 0:00:16.860000
 That is really what the list of topics
 are is a straight demonstration.

0:00:16.860000 --> 0:00:20.900000
 But I'll be taking some time as we
 go through this in particular to go

0:00:20.900000 --> 0:00:25.820000
 through an example of a Kubernetes
 deployment file.

0:00:25.820000 --> 0:00:29.580000
 And other than the Kubernetes deployment
 file, as you'll see, deploying

0:00:29.580000 --> 0:00:34.320000
 a solution into Kubernetes, the deployment
 process itself is really, really

0:00:34.320000 --> 0:00:38.380000
 simple. So let's go ahead and
 let's take a look at this.

0:00:38.380000 --> 0:00:44.620000
 I'm going to start out going into what
 I need to do to actually deploy

0:00:44.620000 --> 0:00:46.400000
 things into Kubernetes.

0:00:46.400000 --> 0:00:51.500000
 And this is where having Kubernetes
 in Azure makes life really pretty

0:00:51.500000 --> 0:00:59.540000
 simple. Okay, Kubernetes has its own,
 let's see if I can bump up the font

0:00:59.540000 --> 0:01:00.220000
 here a little bit.

0:01:00.220000 --> 0:01:03.160000
 There we go, make it a
 little more visible.

0:01:03.160000 --> 0:01:09.980000
 Okay, Kubernetes has its own CLI, a cube
 control, and you can deploy that,

0:01:09.980000 --> 0:01:12.880000
 but you need to authenticate
 into Kubernetes.

0:01:12.880000 --> 0:01:21.760000
 The really cool thing is, if it is an AKS,
 I've got this Azure CLI statement,

0:01:21.760000 --> 0:01:23.560000
 as long as it says, okay, let me go
 ahead and download the credentials

0:01:23.560000 --> 0:01:28.760000
 and use the credentials for that particular
 Kubernetes implementation.

0:01:28.760000 --> 0:01:33.220000
 So it'll log those and that'll be the
 one that your Kubernetes commands

0:01:33.220000 --> 0:01:38.060000
 go again. So I'm going to take that,
 I'm going to copy that, and I'm going

0:01:38.060000 --> 0:01:42.720000
 to go over here to PowerShell,
 and I'm going to run that.

0:01:42.720000 --> 0:01:46.620000
 And that's it right there.

0:01:46.620000 --> 0:01:48.880000
 So it merged that as
 the current context.

0:01:48.880000 --> 0:01:54.260000
 And so now, if I wanted to do something
 such as, let's see if I get this

0:01:54.260000 --> 0:02:05.620000
 right here, kubectlgetnode, should
 give me a list of nodes, and those

0:02:05.620000 --> 0:02:11.040000
 are the nodes that are running for
 that AKS that I have deployed.

0:02:11.040000 --> 0:02:17.840000
 And I can do other things, and this
 is not a Kubernetes CLI course, but

0:02:17.840000 --> 0:02:22.680000
 the point is that I have with really
 little to no effort, provisioned

0:02:22.680000 --> 0:02:27.680000
 in AKS solution service, and I can
 now interact with it, right?

0:02:27.680000 --> 0:02:33.920000
 And I can see information and control it
 simply by downloading the connection

0:02:33.920000 --> 0:02:37.980000
 credentials, and I can
 do that automatically.

0:02:37.980000 --> 0:02:41.160000
 So it's very cool stuff, makes
 your life a lot easier.

0:02:41.160000 --> 0:02:45.280000
 All right, now, what I want to do is
 I'm going to take a look at how you

0:02:45.280000 --> 0:02:48.800000
 will typically deploy solutions
 into Kubernetes.

0:02:48.800000 --> 0:02:52.920000
 Now, I should tell you that there are
 some user interfaces that you can

0:02:52.920000 --> 0:02:54.160000
 run on top of Kubernetes.

0:02:54.160000 --> 0:02:57.180000
 I'll give you a web-based interface
 for Kubernetes, but I'm going to go

0:02:57.180000 --> 0:03:01.640000
 down to what is most common when it
 comes to the point into Kubernetes,

0:03:01.640000 --> 0:03:04.540000
 and that is to use a YAML file.

0:03:04.540000 --> 0:03:11.520000
 YAML is a really very pared-down data
 standard that is entirely built

0:03:11.520000 --> 0:03:14.100000
 around indentation.

0:03:14.100000 --> 0:03:18.340000
 And so, for example, I've ever worked
 with Python, YAML is going to be

0:03:18.340000 --> 0:03:22.480000
 fine for you. But unlike Python, which
 is a programming language, YAML

0:03:22.480000 --> 0:03:28.880000
 is simply a really declarative text standard,
 and it lets you define really

0:03:28.880000 --> 0:03:30.520000
 a hierarchy of settings.

0:03:30.520000 --> 0:03:34.120000
 And in this case, that hierarchy of
 settings is going to represent what

0:03:34.120000 --> 0:03:36.120000
 I want to deploy into Kubernetes.

0:03:36.120000 --> 0:03:40.280000
 And this is by no means going to be an
 exhaustive coverage of a Kubernetes

0:03:40.280000 --> 0:03:44.600000
 YAML file. The file that I'm going to
 have, honestly, I took from an example

0:03:44.600000 --> 0:03:50.460000
 that was on the docs.markself.com site,
 and I've modified it so that it

0:03:50.460000 --> 0:03:53.060000
 matches what I actually
 want to deploy.

0:03:53.060000 --> 0:03:57.360000
 I've got two images.

0:03:57.360000 --> 0:04:02.580000
 I've got a web app image, and I've
 got a background process image, and

0:04:02.580000 --> 0:04:07.700000
 I have both of those as containers,
 and I want to deploy those into a

0:04:07.700000 --> 0:04:09.380000
 Kubernetes solution.

0:04:09.380000 --> 0:04:13.240000
 So let's take a look
 at that process.

0:04:13.240000 --> 0:04:19.680000
 So here is my Kubernetes file, and
 I wasn't all the way up at the top.

0:04:19.680000 --> 0:04:23.920000
 With the Kubernetes file, you've
 got different deployments.

0:04:23.920000 --> 0:04:26.720000
 Now there's different kinds,
 and you'll see that.

0:04:26.720000 --> 0:04:33.200000
 This is a deployment, which is telling
 me is deprecated by V1 beta 2,

0:04:33.200000 --> 0:04:36.440000
 but I'm fine with
 V1 beta 1 for now.

0:04:36.440000 --> 0:04:38.480000
 Now I can also deploy a pod.

0:04:38.480000 --> 0:04:41.840000
 And a pod would be typically
 associated with a container.

0:04:41.840000 --> 0:04:45.860000
 However, even if you really are just going
 to deploy one pod, you typically

0:04:45.860000 --> 0:04:49.360000
 want to use a deployment instead
 for a number of reasons.

0:04:49.360000 --> 0:04:53.100000
 The, to me, greatest reason is that you
 can specify the number of replicas,

0:04:53.100000 --> 0:04:54.780000
 and you can change the
 number of replicas.

0:04:54.780000 --> 0:04:58.160000
 So how many instances of your
 container do you want to run?

0:04:58.160000 --> 0:05:01.720000
 So I've got the API version, the
 kind, which is deployment.

0:05:01.720000 --> 0:05:03.880000
 I've got metadata, which
 is not really that.

0:05:03.880000 --> 0:05:04.600000
 Well, in this case, it is.

0:05:04.600000 --> 0:05:06.180000
 I've got the name
 of the deployment.

0:05:06.180000 --> 0:05:09.000000
 And then I've got the
 specifications.

0:05:09.000000 --> 0:05:10.680000
 It's pretty easy to follow.

0:05:10.680000 --> 0:05:12.120000
 I'm one replica.

0:05:12.120000 --> 0:05:17.380000
 And then the template that I want to
 use, I've got just a label for that.

0:05:17.380000 --> 0:05:20.580000
 The specs for the template.

0:05:20.580000 --> 0:05:23.100000
 So I've got the spec of the deployment,
 and then I've got the spec of

0:05:23.100000 --> 0:05:24.920000
 the template inside of that.

0:05:24.920000 --> 0:05:29.020000
 And you'll notice, again, there's indentation,
 and I'm in VS code right

0:05:29.020000 --> 0:05:32.960000
 now. And so VS code picks that up and
 lets me do things like collapse

0:05:32.960000 --> 0:05:38.660000
 and expand. And so this is going
 to be beta Kubernetes IOS Linux.

0:05:38.660000 --> 0:05:40.780000
 And then it's going
 to have a container.

0:05:40.780000 --> 0:05:43.020000
 I'm going to name this
 container service.

0:05:43.020000 --> 0:05:50.820000
 I'm going to base it on an image that
 is in my Azure container registry.

0:05:50.820000 --> 0:05:55.260000
 And I've already given this Kubernetes
 service access to that container

0:05:55.260000 --> 0:05:58.160000
 registry in order to be
 able to pull images.

0:05:58.160000 --> 0:06:00.480000
 So it should be able to pull
 that image no problem.

0:06:00.480000 --> 0:06:04.500000
 Now, this has resources.

0:06:04.500000 --> 0:06:08.160000
 And this is something, and I don't even
 have this complete, you will typically

0:06:08.160000 --> 0:06:14.000000
 want to limit your resources,
 both CPU and memory, right?

0:06:14.000000 --> 0:06:20.580000
 Because otherwise you could kind of exceed
 what's available and have containers

0:06:20.580000 --> 0:06:24.420000
 have to wait. And then I can also
 specify environment variables.

0:06:24.420000 --> 0:06:31.000000
 The container that I'm deploying is based
 on ASP net core MVC application.

0:06:31.000000 --> 0:06:34.460000
 It's got some configuration options that
 it's going to pull from its standard

0:06:34.460000 --> 0:06:38.240000
 configuration, which in the case of
 running on top of Linux, actually

0:06:38.240000 --> 0:06:40.760000
 Windows as well, can pick
 up environment variables.

0:06:40.760000 --> 0:06:44.920000
 So I'm basically passing some configuration
 to my application.

0:06:44.920000 --> 0:06:47.080000
 Then I've got another deployment.

0:06:47.080000 --> 0:06:49.420000
 The first deployment is a service.


0:06:49.420000 --> 0:06:53.640000
 My second deployment
 is actually my UI.

0:06:53.640000 --> 0:06:55.940000
 And so really kind
 of very similar.

0:06:55.940000 --> 0:07:00.280000
 One thing that's different here, I'm
 specifying the container port.

0:07:00.280000 --> 0:07:05.540000
 So I want this to actually
 have an exposed port 80.

0:07:05.540000 --> 0:07:09.300000
 And then we'll go down through
 here, environment variables.

0:07:09.300000 --> 0:07:11.040000
 And then I have a
 third deployment.

0:07:11.040000 --> 0:07:14.400000
 This time though, the
 kind is service.

0:07:14.400000 --> 0:07:19.040000
 And you can think of a service as really
 giving you an endpoint for your

0:07:19.040000 --> 0:07:22.440000
 containers. A little more complex than
 that, but that will give you an

0:07:22.440000 --> 0:07:24.120000
 endpoint for your container.

0:07:24.120000 --> 0:07:28.100000
 And let's see if it actually
 services a named abstraction.

0:07:28.100000 --> 0:07:31.520000
 It actually gives you this consisting
 of a local port that proxy listens

0:07:31.520000 --> 0:07:36.620000
 on. And the select determines which
 pods will answer requests.

0:07:36.620000 --> 0:07:40.740000
 So the spec for this, it's going
 to be a standard load balancer.

0:07:40.740000 --> 0:07:42.120000
 It's going to listen on port 80.

0:07:42.120000 --> 0:07:51.140000
 And it's going to then hand off to
 a pod named I-N-E-Load UI, which is

0:07:51.140000 --> 0:07:52.760000
 this pod right up here.

0:07:52.760000 --> 0:07:57.940000
 So that is my actual yaml file.

0:07:57.940000 --> 0:08:01.420000
 Now as far as applying the yaml
 file, it's really simple.

0:08:01.420000 --> 0:08:04.520000
 I just have to go and run kubectl.


0:08:04.520000 --> 0:08:05.660000
 And I'm going to apply.

0:08:05.660000 --> 0:08:08.280000
 And the file I'm going to
 apply is this demo file.

0:08:08.280000 --> 0:08:11.880000
 So I'm going to copy that.

0:08:11.880000 --> 0:08:14.500000
 And pop over here.

0:08:14.500000 --> 0:08:19.020000
 First of all, let's make
 sure that file is here.

0:08:19.020000 --> 0:08:20.900000
 I believe it is.

0:08:20.900000 --> 0:08:24.860000
 And there's my I-N-E-Demo yaml.

0:08:24.860000 --> 0:08:25.960000
 I-N-E-Demo yaml.

0:08:25.960000 --> 0:08:28.840000
 There we go. And I apply that.

0:08:28.840000 --> 0:08:31.380000
 All right. Notice how quickly
 that was applied.

0:08:31.380000 --> 0:08:32.240000
 That's pretty awesome.

0:08:32.240000 --> 0:08:35.820000
 However, that doesn't mean that
 it's all actually there.

0:08:35.820000 --> 0:08:49.140000
 OK. So if I go kubectl get and I want
 to get deployment, I can see I've

0:08:49.140000 --> 0:08:50.620000
 got two deployments.

0:08:50.620000 --> 0:08:54.600000
 Now those deployments
 created pods.

0:08:54.600000 --> 0:08:58.540000
 All right. So I've got two pods.

0:08:58.540000 --> 0:09:00.340000
 Those pods are running.

0:09:00.340000 --> 0:09:07.480000
 And then I also have the service
 that got deployed.

0:09:07.480000 --> 0:09:09.620000
 Now, one thing with the service.

0:09:09.620000 --> 0:09:13.440000
 Notice that it says pending
 under external ID.

0:09:13.440000 --> 0:09:16.120000
 So I've got the I-N
-E-Load UI service.

0:09:16.120000 --> 0:09:17.820000
 And over here it says external
 ID is pending.

0:09:17.820000 --> 0:09:22.680000
 So it has not yet been assigned
 an external ID.

0:09:22.680000 --> 0:09:25.700000
 And so I can go and
 get that again.

0:09:25.700000 --> 0:09:31.600000
 And I can also, I believe
 it's, I just forgot.

0:09:31.600000 --> 0:09:34.980000
 I will cheat. Help.

0:09:34.980000 --> 0:09:36.680000
 Throw the help out there.

0:09:36.680000 --> 0:09:40.000000
 I can't remember if it's
 wait or monitor.

0:09:40.000000 --> 0:09:44.600000
 Field file name.

0:09:44.600000 --> 0:09:48.560000
 By the time I apologize,
 by the time I find this.

0:09:48.560000 --> 0:09:56.000000
 OK. Watch is what it is.

0:09:56.000000 --> 0:09:59.400000
 OK. We're going to go with watch.

0:09:59.400000 --> 0:10:01.000000
 Which again, it's probably
 took enough time.

0:10:01.000000 --> 0:10:01.960000
 That's probably already there.

0:10:01.960000 --> 0:10:03.440000
 No, but it's still pending.

0:10:03.440000 --> 0:10:09.140000
 All right. So I'm going to watch this
 until that public IP gets assigned.

0:10:09.140000 --> 0:10:10.900000
 And it'll show up automatically.

0:10:10.900000 --> 0:10:14.500000
 All right. And now I
 have an external IP.

0:10:14.500000 --> 0:10:20.460000
 I can see this 1332,
 or 1392, 115, 135.

0:10:20.460000 --> 0:10:23.380000
 I'm going to go ahead and hit that
 and control C out of this, because

0:10:23.380000 --> 0:10:27.280000
 I don't need that anymore.

0:10:27.280000 --> 0:10:32.340000
 And now, the moment of truth.

0:10:32.340000 --> 0:10:36.460000
 Apparently I did not
 actually copy it.

0:10:36.460000 --> 0:10:38.440000
 That would be the first
 mistake there.

0:10:38.440000 --> 0:10:41.080000
 Let's try that again.

0:10:41.080000 --> 0:10:47.320000
 Always good to remember how to use
 the tools that you're working with.

0:10:47.320000 --> 0:10:50.760000
 There we go. There is my web app.

0:10:50.760000 --> 0:10:58.400000
 And should be working properly, I believe,
 that I can even annoy my service.

0:10:58.400000 --> 0:10:59.900000
 That may take a moment.

0:10:59.900000 --> 0:11:01.720000
 Not sure I fully set that one up.

0:11:01.720000 --> 0:11:07.380000
 There we go. I ping the service and it
 came back to me with a sassy reply

0:11:07.380000 --> 0:11:10.880000
 there. But I've got a fully
 functional system.

0:11:10.880000 --> 0:11:15.960000
 It means that not only did the container
 with my web app work, but I've

0:11:15.960000 --> 0:11:20.020000
 deployed a web app, I've deployed a
 service that's monitoring the queue.

0:11:20.020000 --> 0:11:23.560000
 The way that little interchange work
 is my web app drops a message on

0:11:23.560000 --> 0:11:27.760000
 a queue. The server is the service
 is monitoring the queue.

0:11:27.760000 --> 0:11:32.160000
 It pulls that message off and then puts
 a message into another queue that

0:11:32.160000 --> 0:11:33.420000
 the web app is looking for.

0:11:33.420000 --> 0:11:35.640000
 So kind of a standard queue
-based communication.

0:11:35.640000 --> 0:11:40.380000
 But the point is not that I can get
 snarky answers from a service.

0:11:40.380000 --> 0:11:45.480000
 The point is that I now really have a
 fully functional, if somewhat useless

0:11:45.480000 --> 0:11:47.800000
 solution in Kubernetes.

0:11:47.800000 --> 0:11:50.120000
 And before we go, I want to show you
 a couple other things, because this

0:11:50.120000 --> 0:11:52.540000
 is pretty cool with Kubernetes.

0:11:52.540000 --> 0:11:58.800000
 Now, if I go to the resource group where
 I have dropped Kubernetes, where

0:11:58.800000 --> 0:12:01.840000
 I created Kubernetes, the only thing
 that I see in this resource group

0:12:01.840000 --> 0:12:11.780000
 is my service. However, when you provision
 a Kubernetes service, it actually

0:12:11.780000 --> 0:12:15.280000
 creates a secondary
 resource group.

0:12:15.280000 --> 0:12:17.880000
 And it gives itself rights
 to that resource group.

0:12:17.880000 --> 0:12:20.520000
 And if I look at what I have,
 I've got an availability set.

0:12:20.520000 --> 0:12:22.720000
 I've got some disks.

0:12:22.720000 --> 0:12:24.580000
 I've got a load balancer.

0:12:24.580000 --> 0:12:26.340000
 I've got network interfaces.

0:12:26.340000 --> 0:12:30.700000
 I've got a network security group,
 IP address for the load balancer.

0:12:30.700000 --> 0:12:35.800000
 I've got built-in route table, virtual
 machines, and the virtual network.

0:12:35.800000 --> 0:12:42.480000
 And if I go, for example, to the load balancer,
 and if I go to load balancing

0:12:42.480000 --> 0:12:49.340000
 rules, I will see that this should
 be the load balancing rule for my,

0:12:49.340000 --> 0:12:52.480000
 for my 80. That's my port
 80 load balancing rule.

0:12:52.480000 --> 0:12:54.540000
 It's going to drop port 80.

0:12:54.540000 --> 0:12:56.640000
 It's just pushing port
 80 back to those VMs.

0:12:56.640000 --> 0:12:58.820000
 It doesn't really care
 what the VMs did.

0:12:58.820000 --> 0:13:03.980000
 Likewise, I have a relatively restrictive
 network security group.

0:13:03.980000 --> 0:13:09.220000
 The cool thing is, when I pop over
 there, I have this inbound security

0:13:09.220000 --> 0:13:10.600000
 rule for port 80.

0:13:10.600000 --> 0:13:14.320000
 So Kubernetes automatically
 added this rule.

0:13:14.320000 --> 0:13:19.260000
 It's managing all of this
 infrastructure for me.

0:13:19.260000 --> 0:13:22.220000
 And it's automatically provisioned
 three virtual machines.

0:13:22.220000 --> 0:13:25.100000
 It's put those three virtual machines
 in an availability set.

0:13:25.100000 --> 0:13:26.660000
 So I've got high availability.

0:13:26.660000 --> 0:13:28.420000
 I've got load balancing.

0:13:28.420000 --> 0:13:33.980000
 And all I need to be able to do is,
 I need to be able to define what it

0:13:33.980000 --> 0:13:38.740000
 is I want to deploy in a yaml
 file and then deploy that.

0:13:38.740000 --> 0:13:42.660000
 And that really, if you think
 about it, is pretty amazing.

0:13:42.660000 --> 0:13:47.080000
 The workload that I just provisioned,
 okay, it was one container of two

0:13:47.080000 --> 0:13:49.780000
 different, or one instance of
 two different containers.

0:13:49.780000 --> 0:13:52.960000
 But I could very easily
 modify this.

0:13:52.960000 --> 0:13:57.020000
 And in fact, I could reapply this if I
 wanted to bump it up to three instances

0:13:57.020000 --> 0:14:02.220000
 per deployment. I could do that by simply
 change that number and reapplying

0:14:02.220000 --> 0:14:03.460000
 the same yaml file.

0:14:03.460000 --> 0:14:08.420000
 So a lot of functionality, a lot of
 capability here for what I have to

0:14:08.420000 --> 0:14:12.220000
 say is pretty much the minimum effort
 that you could pretty much do.

0:14:12.220000 --> 0:14:16.740000
 And I hope to have to put out to
 manage a clustered environment.

