A node may be a VM or physical machine, depending on the cluster.

The scheduler distributes work or containers across multiple nodes.

Have a good understanding of YAML files and when to use them:

List, dictionary, List of Dictionaries.

kubectl get pods -o wide

kubectl describe pods <POD_NAME>

kubectl delete pod <POD_NAME> --grace-period=0

kubectl get pod -o yaml

kubectl run redis --image=redis123 --dry-run=client -o yaml > redis.yaml

kubectl get replicaset

kubectl scale replicaset <REPLICASET_NAME> --replicas=2

kubectl delete replicaset <REPLICASET_NAME>

Check for apiVersion: kubectl api-resources | grep replicaset

kubectl scale replicaset new-replica-set --replicas=0

Rollout and Versioning

Recreate (Face downtime) vs RollingUpdate (Take down one by one)

Summarize Commands

• kubectl create –f <DEPLOYMENT_YAML_FILE>

• kubectl get deployments

• kubectl apply –f <DEPLOYMENT_YAML_FILE>

• kubectl set image deployment/<DEPLOYMENT_NAME> <CONTAINER_NAME>=<NEW_IMAGE>

• kubectl rollout status deployment/<DEPLOYMENT_NAME>

• kubectl rollout history deployment/<DEPLOYMENT_NAME>

• kubectl rollout undo deployment/<DEPLOYMENT_NAME>

Enable connectivity between a group of pods and an external source

Node port, Cluster IP, Load Balancer

Node Port:

• Port range: 30000 to 32767
• minikube service myapp-svc --url

Cluster IP:

• Default Setting
• 

Load Balancer:

• Works as expected only with Cloud Providers like GCloud, AWS, Azure.
• Otherwise act as Nodeport.