Ladies and gentlemen, we are now at part 3, which is the last part of the Docker & Kubernetes series. Now that we know what Docker/Kubernetes is, what containers are, and we have minikube installed, let's move on to deployments and pods.
First and foremost, what are deployments and pods?
Deployments are a way to bring up your environment from your Docker images. They allow you to have your golden environment, with as many pods as you want, with the ability to update those pods on the fly and give you self-healing.
Pods Pods are a collection of containers. You can have multiple pods, with multiple containers inside of those pods. It also allows you to manage storage resources, unique network IPs, and options that show how a container should run.
Let's ensure that our minikube node is up and operational. Run the following and you should see the output below:
Now, let's create a Kubernetes manifest which will be stored locally (in production, you always want to store these in some sort of private source control). Manifests are written in YAML, and we can do a vim to edit the code, or use VSCode
I have written the following manifest, which will spin up an Nginx deployment with 5 replicas (pods).
Lets break this down:
1) The API version is the version of the Kubernetes API that you will be interacting with.
2) The Kind is for specifying the type of manifest. In our case, it's Deployment.
3) The metadata is, well.. the metadata :). This is the metadata of your deployment.
4) The spec block is for building your pods and specifying what you want them to look like.
Take a look at "replicas". Replicas state how many pods you will be creating.
In our case, we will be using the latest version of Nginx and allowing traffic over port 80.
To kick this off, we will want to use the Kubernetes API.
kubectl create -f TestManfiest.yaml
You should see the following:
deployment.apps "nginx-deployment" created
Now we will run the following to ensure that our pods got created:
kubectl get pods
You should see a similar output to this screen
And like magic, the pods are created!
Next, we want to create a service. A service allows your deployment/pods to be accessible over the web and internally. The definition of a service is "an abstraction which defines a logical set of Pods and a policy by which to access them"
To expose our deployment, we want to run the following:
kubectl expose deployment nginx-deployment --port=80 --type=LoadBalancer
What the above does is creating a service and exposes our deployment over port 80 in a Load Balancer fashion (in short: a load balancer allows you to span application across multiple endpoints vs having a single point of failure).
There you have it! You're up, running, and ready to go with Nginx.