Most time in Kubernetes (aka k8s) we can use the command kubectl to start a deployment.

But in production environment we would better use the declarative configuration such as YAML to define a resource, and start the deployment from that YAML. This is b/c the declarative configuration is easier for version control.

For instance, we can start the deployment using this command,

 $ kubectl run demo --image=cloudnatived/demo:hello --port=8888
 $ kubectl port-forward demo 9999:8888
 $ curl --head http://localhost:9999
 

Instead we would do it better by generating a config file, which is controlled via version systems such as Git.

 $ kubectl run demo --image=cloudnatived/demo:hello --port=8888 --dry-run -o yaml  > deploy.yaml
 $ kubectl apply -f deploy.yaml 
 pod/demo created
 

Hence this is the content in deploy.yaml:

 apiVersion: v1
 kind: Pod
 metadata:
   creationTimestamp: null
   labels:
     run: demo
   name: demo
 spec:
   containers:
   - image: cloudnatived/demo:hello
     name: demo
     ports:
     - containerPort: 8888
     resources: {}
   dnsPolicy: ClusterFirst
   restartPolicy: Always
 status: {}
 

Everyone can change the deployment. But with a good version control, we can avoid many conflicts.