Containers don’t come with the weight of a full OS image and a hypervisor.

Containers include only OS processes and dependencies necessary to execute the code inside them.

Containers occupy less space(MBs not GBs) which means you can run multiple copies of the same application on the same hardware.

Containers are faster and easier to deploy, provision and restart. This make them ideal for CI/CD pipelines and a better fit for Agile development teams.

Docker enhances the native Linux containerization technologies which enables:

• seamless portability: can run natively across any desktop, data center or cloud development
• granular updates: only one process runs in a container. An application can continue running without a container being taken done
• automatic container creation: based on application source code
• container versioning: tracking image versions and roll backs are fast
• container reuse: base images can be used as templates for nee containers
• shared container libraries: an open source user contributed container images registry

DockerFile - every container has one and states the instructions for building the container’s image. It consists in command-line instructions;

Docker images - contains executable source code, libs and dependencies the code needs; Read-only file; A Docker image can be build from scratch, most developers reuse common ones from repositories;

Docker containers - the live and running instance of the Docker image; Ephemeral, users can interact with them, run commands and adjust settings using docker commands

Docker registry - open source distribution system and versioning using git

On the same host:

• Docker Compose - creates a .yaml which specifies the used containers; can deploy and run containers from a single command; can define persistent volumes for storage, databases and service dependencies;

On different hosts:

You need container orchestration tools.

• Docker Swarm
• Kubernetes
• AWS EKS
• OpenShift