Some videos:

• https://www.youtube.com/watch?v=3c-iBn73dDE&t=1s

• Has info on how to use the images
• Includes configuration info, like environmental variables
  • That's what the -e represents

docker pull docker run docker start docker stop docker ps docker exec -it docker logs

docker help

• To see all possible docker commands

docker build

• Turns the file into an image

Excludes the files you don't want to end up in a docker image

• Docker Scout allows you to identify any security vulnerabilities for each layer in an image
• It extracts the software bill of material from the image and compares it to security advisory databases
• When there's a match, it's given a severity rating (to prioritise safety efforts)

docker images

• View the current images loaded on host machine

docker ps

• Returns list of current RUNNING containers
• On Docker desktop, you can click on the running container to inspect the logs, view the file system, or execute commands directly in the running container

docker ps -a

• Returns list of running AND stopped containers

• Download it from the cloud
• If version is not specified, it pulls the latest version

docker pull :

OR: docker run :

• Basically 2 commands in 1: Pulls images AND starts container

eg. docker run redis:4.0

docker run: Creates new container with a command

• Combines "docker pull" & "docker start" in one command -> Pulls image AND starts container docker run <image-name>:<version>

eg. docker run postgres:9.6

docker run -d

• The -d creates container in a detatched mode
• Detached mode allows you to use the terminal again? TBC

• If >=2 containers are listening to the same port, ensure you bind them to different ports on your HOST machine
• You need to create binding between container port and laptop's port for the container to be reachable by any application
• Specify the binding of the ports during the run command
• One laptop's port can only run one container
• -p allows you to bind the port of the host machine to the container docker run -p<laptop's-port>:<container's-port> <image-name> docker run -p6000:6379 redis

docker run -d -p<host-port>:<container's-port> --name <new-container's-name> <name-of-image-you-want>

• The -d is just to run it in detached mode (optional)

docker container stop <container-name>

OR, to forcefully stop it: docker container kill <container-name>

OR, to remove the container entirely: docker container remove <container-name>

docker stop <container-id>
docker start <container-id>

docker logs <container-id> OR docker logs <container-name>

• Check the logs of the container

docker exec -it <container-id> /bin/bash OR docker exec -it <container-name> /bin/bash

• -it: interactive terminal
• Brings you inside the container's terminal as a user

ls OR pwd

• Check which directory you're in
• If it is empty, to exit directory use

/

env

• Print the environmental variables to check if things are set correctly
• Useful if your container has a complex configuration & you want to validate that things are set up correctly
• OR, if you're running your own application that you wrote in a container

exit

• Exits the terminal

• For managing multi-container applications
• Allows you to define multiple applications and their DOcker images in a single YAML config file
• Eg. frontend, backend, and a database
• To run all these containers simultaneously, use: docker compose up
• To stop these running containers, use: docker compose down

• Uploads your image into a remote registry, where it can run on a cloud (eg. AWS) with elastic container service(!!!), or it can be launched on serverless platforms (eg. Google Cloud Run)

• Abstracts away the kernel to make it possible for your hosts to run different docker images

• On a larger scale, orchestration tools like kubernetes (aka K8s) is needed to run and manage containers
• A control plane exposes an API that manages the cluster, which has multiple nodes/ machines
• Each node/machine contains a kublet & multiple pods
• Pod: The minimum deployable unit in kubernetes, and has one/more containers inside
• Kubernetes is effective as it automatically scales according to the desired state of the system you describe
• Has fault tolerance to auto-heal if any server goes down