Convert Pod to Deployment in Kubernetes

Make your application more reliable

Pod to Deployment Conversion

A deployment allows you to run more than 1 replica of the same pod thus, giving you more reliability. You can scale up a Deployment to facilitate and handle more load.

In this brief write-up, I demonstrated how to convert your existing pod to a deployment.

Getting Started

Below is an API service running in the Kubernetes cluster which runs as a single pod.

k8s@terminal:~$ k get pods -n reliable
NAME      READY   STATUS    RESTARTS   AGE
med-api   1/1     Running   0          2m9s

Get the current Pod’s YAML

k8s@terminal:~$ k get pods med-api $do -oyaml -n reliable > single-pod.yaml
vim single-pod.yaml
....
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: med-api
  name: med-api
  namespace: reliable
spec:
  containers:
  - image: nginx
    imagePullPolicy: Always
    name: med-api
    resources: {}

There are multiple ways to do this, one is to copy a Deployment example from https://kubernetes.io/docs and then merge it with the existing Pod YAML.

Let’s demonstrate this: Download a Deployment example manifest from Kubernetes Doc and edit it to match the desired outcome.

Now copy/use a Deployment example YAML and put the Pod’s metadata: and spec: from single-pod.yaml into the Deployment’s template: section:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx-deployment # <--- change this 
labels:
    app: nginx # <--- change this to match label on the pod
spec:
  replicas: 3
  selector:
    matchLabels:
      app: nginx  # <--- change this to match label on the pod
  template: # => from here down copy and paste the pods metadata: and spec: sections
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx:1.14.2
        ports:
        - containerPort: 80

The final Deployment Manifest looks like this….

apiVersion: apps/v1
kind: Deployment
metadata:
  name: med-api
  namespace: reliable
  labels:
    run: med-api
spec:
  replicas: 3
  selector:
    matchLabels:
      run: med-api
  template:
  # => from here down its the same as the pods metadata: and spec: sections
    metadata:
      labels:
        run: med-api
      name: med-api
      namespace: reliable
    spec:
      containers:
      - image: nginx
        imagePullPolicy: Always
        name: med-api
        resources: {}

Now, let’s deploy the merged deployment manifest.

k8s@terminal:~$ k apply -f single-pod.yaml 
deployment.apps/med-api created
k8s@terminal:~$ k get pods -n reliable
NAME                       READY   STATUS    RESTARTS   AGE
med-api                    1/1     Running   0          21m
med-api-67bb7c7577-chng5   1/1     Running   0          8s
med-api-67bb7c7577-khhcx   1/1     Running   0          8s
med-api-67bb7c7577-tccn2   1/1     Running   0          8s
k8s@terminal:~$ k delete pod med-api -n reliable
pod "med-api" deleted
k8s@terminal:~$ k get pods -n reliable
NAME                       READY   STATUS    RESTARTS   AGE
med-api-67bb7c7577-chng5   1/1     Running   0          55s
med-api-67bb7c7577-khhcx   1/1     Running   0          55s
med-api-67bb7c7577-tccn2   1/1     Running   0          55s

Notice that the Deployment has created all three replicas, and all replicas are up-to-date (they contain the latest Pod template) and available.

Reliable k8s workloads make a happy Engineer!!!