Skip to content

Latest commit

 

History

History
70 lines (61 loc) · 2.33 KB

KUBE.md

File metadata and controls

70 lines (61 loc) · 2.33 KB
Raw

Kubernetes notes

Kubectl cheatsheet

  • minikube start

    Local k8s implementation to test k8s locally

  • minikube stop
  • minikube delete

    Sometimes necessary to clear local state before running start again

  • kubectl get nodes
  • kubectl get pods
  • kubectl get deployments
  • kubectl get services
  • kubectl scale deployments/<deployment_name> --replicas=<desired_number>

    Fun command but bot won't receive this kind of traffic

  • kubectl set image deployments/<deployment_name> <deployment_name>=<new_docker_image_name>

    This will perform a rolling update

  • kubectl rollout undo deployments/<deployment_name>

    Undo a rolling update back to previous state

  • minikube dashboard

Demo run

  1. minikube start
  2. kubectl create deployment hello-minikube --image=k8s.gcr.io/echoserver:1.10

Create deployment

  1. kubectl expose deployment hello-minikube --type=NodePort --port=8080

Create service using the newly created deployment

  1. kubectl get pod

Verify pod is launched and ready

  1. minikube service hello-minikube --url
  2. View the url to see status page
  3. kubectl delete services hello-minikube
  4. kubectl delete deployment hello-minikube
  5. minikube stop
  6. minikube delete

To use local docker images

  • minikube docker-env
  • eval $(minikube -p minikube docker-env)

Bot benefits

  • Service recovery if bot goes down
  • Rolling deployments (no downtime)

Clusters

  • Single master node
  • 1 or many Nodes

Nodes

  • The workers in the cluster
  • Runs a Kubelet to communicate with master
  • Runs a container runtime

Pods

  • Logical abstraction that can represent a group of one or more containers
  • Containers in a pod share the same IP address and ports
  • Containers in a pod run in a shared-context
  • In a pod you might run a persistent volume app and web server app together

Deployments

  • A configuration file that master will use to maintain state
  • This is the self-healing part of k8s
  • Node goes down? Machine failure? Master will re-run the config

Services

  • Allows applications to receive traffic
    1. ClusterIP - service is available within the same cluster
    2. NodePort - service is available outside the cluser
    3. LoadBalancer - creates fixed external IP
    4. ExternalName - exposes service by name (CNAME record)
  • Services are exposed Deployments meaning traffic can come through