API Service

We've deployed one service, and we've deployed multiple instances of it. Time to deploy a second service!

This service doesn't serve a webpage! It's a JSON API. It's the backend for our chat application. By deploying the API and configuring the front-end to talk to it, we'll have a functional chat application!

Create a Deployment Configuration

Let's write a deployment from scratch.

Feel free to reference the k8s docs here as you go for examples of the proper structure.

Create a new file called api-deployment.yaml.
Add the apiVersion and kind fields. The apiVersion is apps/v1 and, since this is a deployment, the kind is Deployment.
Add a metadata/name field, and let's name our deployment synergychat-api for consistency.
Add a metadata/labels/app field, and also set it to synergychat-api. This will be used to select the pods that this deployment manages.
Add a spec/replicas field and let's set it to 1. We can always scale up to more pods later.
Add a spec/selector/matchLabels/app field and set it to synergychat-api. This should match the label we set in step 4.
Add a spec/template/metadata/labels/app field and set it to synergychat-api. Again, this should match the label we set in step 4. Labels are important because they're how Kubernetes knows which pods belong to which deployments.
Add a spec/template/spec/containers field. This actually contains a list of containers that will be deployed:
1. Note: A hyphen is how you denote a list item in YAML
2. Set the name of the container to synergychat-api.
3. Set the image to bootdotdev/synergychat-api:latest. This tells k8s where to download the Docker image from.

I don't want to give you the YAML because I want you to type it out, but here's the equivalent JSON:

{
  "apiVersion": "apps/v1",
  "kind": "Deployment",
  "metadata": {
    "name": "synergychat-api",
    "labels": {
      "app": "synergychat-api"
    }
  },
  "spec": {
    "replicas": 1,
    "selector": {
      "matchLabels": {
        "app": "synergychat-api"
      }
    },
    "template": {
      "metadata": {
        "labels": {
          "app": "synergychat-api"
        }
      },
      "spec": {
        "containers": [
          {
            "name": "synergychat-api",
            "image": "bootdotdev/synergychat-api:latest"
          }
        ]
      }
    }
  }
}

Create the Deployment

kubectl apply -f api-deployment.yaml

Next, take a look at all the pods you have running now. You should see pods for the web service and a pod for the api service.

However, you might notice that the api pod isn't in a "ready" state. In fact, it should be stuck in a "CrashLoopBackOff" status. Oh no! We've created a thrashing pod!

Run:

kubectl proxy

Run and submit the CLI tests.