containers

notes from working through the 544 tutorial

Prerequisite: Set up a cloud VM (Azure) as a workstation
- Uses VSCode to ssh to a bash shell

Docker is a software development platform and a virtualization technology; key concept is hermetic operation, no shock waves into the system at large
- except the shock waves we want of preserving compute module results
Docker applications can therefore run anywhere; they are system agnostic
- Docker communicates “natively” (directly) with the system kernel
Resource optimization: Multiple Docker containers run on a single system (VM say)…
- …from a common source image? Share files to reduce footprint!
Resource provisioning
- Docker Container assigned so much RAM, CPU cores etcetera by a management service
- For example on AWS use Elastic Container Service to construct a Task…
  - …Task: Composed of many Container runs
Bottom line of Containers is Run Many, Run Anywhere: Scale and consistency
- VMs can be bulky and slow
  - Containers only virtualize the OS, not the hardware
Three fundamental elements: Docker file, Docker image, Docker container
- Docker File is like DNA: Code: How to build an image
  - Start with FROM base_image
  - RUN commands to customize
  - ENV commands to set environment variables used inside the running Container
  - CMD or ENTRYPOINT to stipulate what happens at runtime
- Docker Image: an immutable file
  - A snapshot of software + dependencies all the way down to the OS level
  - Configured content can be spun up as a container
- Docker Container is the executable compute module

a dockerfile resides in a host folder
a Docker image lives somewhere (not our concern, managed behind the scenes)
the docker command is a command line interface
- interfaces with the docker daemon/engine that manages everything docker
On a PC: Enabled by Docker Desktop (app) + WSL2