There are 2 ways to create a kubernetes object
	a. Ad-hoc using kubectl create
	b. Declarative approach using a file.
		kubectl apply|create -f FILE

Namespaces
==========
1. Create two namespace: testing and dummy
kubectl create namespace testing
kubectl create ns dummy

2. List all the namespaces
kubectl get ns

3. Find out which is the current namespace for the user.
kubectl config get-contexts

4. Change it to testing namespace
kubectl config set-context --current --namespace testing
kubectl config get-contexts

5. Try listing pods from kube-system namespace. Why can't you see the pods in your namespace?
kubectl get po
kubectl get po -n kube-system

6. Delete the dummy namespace.
kubectl delete ns dummy
kubectl get ns

Pods
====
1. Create a pod using the nginx image.
kubectl run web --image=nginx

2. Display the pod.
kubectl get po -o wide
kubectl get po -o custom-columns=NAME:.metadata.name,NODE:.spec.nodeName,IP:.status.podIP
NAME	NODE	IP
web 	worker	10.20.30.40

3. Try to visit the pods webpage from the other node.
curl 10.20.30.40

4. Try using the worker node
curl worker

5. Quiz: Why did it work for step 3 but not 4? Discuss with instructor.

6. Delete the pod
kubectl delete po web

Jobs
====
We will be creating the job using the declarative approach.
1. Create a job that will end in a few seconds. make sure the conntainer name is lazy.
kubectl create job short --image=busybox --dry-run=client -o yaml -- sleep 3 > short.yaml
vi short.yaml
...
spec:
  template:
    spec:
      containers:
      - command:
        - sleep
        - "3"
        image: busybox
        name: lazy		# Change this line
...
kubectl create -f short.yaml

2. Check the status of the job and pod
kubectl describe job short
kubectl get job,po

3. Set the job to run 5 times and 2 in parallel each time. Make sure the job does not take 15 seconds to finish.
vi short.yaml
...
kind: Job
  name: short
spec:
  completions: 5		# Add this line
  parallelism: 2		# Add this line
  activeDealineSeconds: 15	# Add this line
  template:
    spec:
...

4. Open another terminal and monitor live events.
kubectl get events -w

4. Back in the first terminal, recreate the job.
kubectl delete job short
kubectl apply -f short.yaml

5. Monitor the events in the second terminal and type the following commands in the first terminal
kubectl get job,po

6. Clean up. Delete the job and exit the live event monitoring.
kubectl delete job short
Ctrl + C	# exit the kubectl get events -w

CronJobs
========
1. Create a cronjob using the 3s-cronjob.yaml file. Correct any errors encountered in the file.
kubectl apply -f 3s-cronjob.yaml

2. Monitor the job. The job will only run after every 2 minutes.
kubectl describe cronjob repeat
kubectl get cj,job,po 

3. Clean up. Delete the CronJob.
kubectl delete cj repeat