Docker is for “how to run one container”.

K8s is for “how to manage/run lots of containers”.

Docker

Docker is for “how to run one container”

Docker components

Installation on macOS

Goto docker website

Practices with the “Docker Desktop”

Practices with the CLI

About the images

# list Images
docker Images

# pull the latest image
docker pull ubuntu

# pull 'ubuntu' image with specific version 'jammy'
docker pull ubuntu:jammy

# remove an image
docker rmi ubuntu

About the containers

# list the running containers
docker ps

# show all the containers
docker ps -a

Create and run containers.

# create and run a container (with random name) from an image
docker run ubuntu

# create and run a container with specific name `ubuntu-container` from image ubuntu
docker run --name ubuntu-container ubuntu

# create and run a container with interactive mode
# --name ubuntu : container name
# -i : interactive Keep STDIN open even if not attached
# -t : Allocate a pseudo-TTY
docker run --name ubuntu -i -t ubuntu /bin/bash

# run a welcome-to-docker container and open the port 8080,
# so can access it from browser
# open a browser, open url `http://localhost:8080`
docker run -d -p 8080:80 docker/welcome-to-docker

Stop running containers

docker kill ubuntu

Remove containers

# remove the container ""
docker rm ubuntu-container
# or use the container ID which can be seen from the `docker ps -a`
docker rm 740ac834e823

Start existing containers

# start an existing container
docker start ubuntu

# Attach local standard input, output, and error streams to a running container
docker attach ubuntu

# display the logs of a container
docker logs ubuntu

# display the running processes of a container
docker top ubuntu

Composing the image

cd ~/ws/demos/docker
mkdir static_web
cd static_web
touch Dockerfile

Here’s the content of the Dockerfile

# Version: 0.0.4
FROM ubuntu:24.04
LABEL maintainer="peter.2.song@outlook.com"
RUN apt update && apt install -y nginx
#RUN echo 'Hi, I am in your container' > /usr/share/nginx/html/index.html
RUN echo 'Hi, I am in your container' > /var/www/html/index.html
EXPOSE 80
# start the nginx when using `docker run`
# use ENTRYPOINT to start nginx
ENTRYPOINT ["/usr/sbin/nginx", "-g", "daemon off;"]
# use CMD to start nginx
#CMD nginx -g "daemon off;"

# build an image from the Dockerfile above
docker build -t "petesong/static_web" .

# run a container from the image
# if there's ENTRYPOINT in the Dockerfile, can run a container like this
docker run -d -p 8081:80 --name static_web petesong/static_web

# 访问
curl localhost:8081

Kubernetes(K8s)

K8s is for “how to manage/run lots of containers”

Core functions: 核心功能

Kubernetes 架构

Installation on a local computer

https://kubernetes.io/docs/tasks/tools/

Practices with minikube on macOS

Install minikube

Install minikube per link , the kubectl would also be installed.

brew install minikube

kubectl version
minikube version

Start minikube and the cluster

Start the cluster after have started the docker desktop

minikube start --driver=docker

kubectl cluster-info

kubectl cluster-info dump

Minikube addons

# list addons
minikube addons list

# enable an addon
minikube addons enable metrics-server

# disable an addon
minikube addons disable metrics-server

Interact with the clusters

Interact with the clusters:

Use minikube dashboard

minikube addons enable metrics-server

minikube dashboard

The dashboard should be opened in the browser automatically.

Use kubectl CLI

Use kubectl command line to interact with the cluster

# create a deployment
kubectl create deployment hello-node --image=registry.k8s.io/e2e-test-images/agnhost:2.39 -- /agnhost netexec --http-port=8080

# scale up a deployment
kubectl get deployments,replicasets,pods
kubectl scale deployments/hello-node --replicas=3
kubectl get deployments,replicasets,pods

# scale down a deployment
kubectl get deployments,replicasets,pods
kubectl scale deployments/hello-node --replicas=1
kubectl get deployments,replicasets,pods


# upgrade the version of the app
# first get the container name, e.g. `agnhost`
kubectl get deployment hello-node -o wide
# update the version
kubectl set image deployments/hello-node agnhost=registry.k8s.io/e2e-test-images/agnhost:2.40
# check the pods, can see one pod is terminating
kubectl get pods
# wait for a while, check the deployment
# can see the new version of app
kubectl describe deployment hello-node


# delete deployment
kubectl get deployments
kubectl delete deployment hello-node
kubectl get deployments

Want to see the resource types

# list all the resource types
kubectl api-resources

# list the deployment
kubectl get deployments

# list the nodes
kubectl get nodes

# list the pods
kubectl get pods
# show detailed information
kubectl describe pods

# list the events to see the
kubectl get events

Want to work with services

# by default, the Pod is only accessible by its internal IP address within the Kubernets cluster.
# to make the hello-node container accessible from outside the Kubernets virtual network
# you have to expose the Pods as a Kubernets service
kubectl expose deployment hello-node --type=LoadBalancer --port=8080

# get services
kubectl get services

# use the service
# it opens up a browser window the serves your app and show the app's response
minikube service hello-node
# alternatively, we can use the `kubectl` to forward the port
# then open http://localhost:7080 in the browser
kubectl port-forward service/hello-node 7080:8080


# delete service
kubectl get services
kubectl delete service hello-node
kubectl get services


# get logs of the specific pod with the pod name showed above.
kubectl logs hello-node-66d457cb86-g6xcb

# execute command on the container
# list the environment variables
kubectl exec hello-node-66d457cb86-g6xcb -- env
# run bash in the container
kubectl exec -ti hello-node-66d457cb86-g6xcb -- bash

# check the resources
kubectl top pods
kubectl top nodes

# view the config
kubectl config view

Use the curl to interact with APIs

# by defaut, the pods are running inside Kubernets are running on a private, isolate networks.
# we need to open an second terminal window to run the proxy that will forward communications
# into the cluster-wide, private network
kuberctl proxy

then

curl http://localhost:8001/version

# get the pod name
export POD_NAME=$(kubectl get pods -o go-template --template '\n')

echo $POD_NAME

# access the Pod through the proxied API
curl http://localhost:8001/api/v1/namespaces/default/pods/$POD_NAME:8080/proxy/

Use the curl to use the service

# it opens up a browser window the serves your app and show the app's response
minikube service hello-node --url
# can see the url, e.g. http://127.0.0.1:58210
# use the url above
curl http://127.0.0.1:58210

# alternatively, we can use the `kubectl` to forward the port
# then open http://localhost:7080 in the browser
kubectl port-forward service/hello-node 7080:8080
# then
curl http://localhost:7080

Run bash in the container

# run bash in the container
kubectl exec -ti hello-node-66d457cb86-g6xcb -- bash
# then we can use `curl` inside the container
curl http://localhost:8080
# then exit the container
exit

Stop minikube and the cluster

minikube stop

References

For K8s:

For Docker: