Overview

Docker makes software dev and deploying applications much easier by packaging applications w/ all necessary dependencies, configurations and system tools and runtime into a container
Creating a standardized unit that has everything the application needs to run
Containers allow us to NOT install any of the services in your own machine
- We just run that server as a docker container using a single command — docker run postgres, for example
- If you’re using 10 services on which your application depends on, that would mean just running 10 containers → Docker run postgres, docker run redis, … etc.
- You can also have different versions of the same application running
Deployment w/o docker → Installations and configurations need to be done on the server itself, can have different versions installed,… lots of conflicts arise during setup
Deployment w/ docker → Docker artifact gets packaged with the application code, but also the configurations and dependencies → NO configurations / manually installation required on server
- All we need to do on the server now is have docker runtime installed to be able to run any containers

Virtual Machine vs Docker

Understanding how an Operating System (OS) is made up

There are two main components / layers
- OS Kernal → Core of every OS
  - Communicates with the hardware (CPU, memory, etc) and the software
  - Middle man between the hardware and the software
- OS Application Layer → Software apps such as google chrome, MS word, etc
Docker virtualizes the OS Application Layer, while using the host kernal
- Applications are installed on top of that docker application layer now
VM virtualizes the application layer AND the OS Kernal.
- This means that when you download a VM, it comes with it’s own OS, which runs on top of the CPU OS!
Sizes of the docker packages / images are much smaller as they don’t have to bring in their own Operating Systems!
Docker containers are much faster to start (uses host OS, and just boots up the application layer on top of it )
VM is better in terms of compatibility → you can run the VM on ANY OS, but docker can’t run linux apps on a windows OS
- This is one of the cons of docker, but docker made a docker desktop app which uses a hypervisor layer with a lightweight linux distro which allows windows and mac OS to run docker apps
- This means we would install docker desktop to run linux based images! (Which is the industry common practice, as most DBs, apps use this)

Installation

Install from here:

https://www.docker.com/products/docker-desktop/
This comes with Docker engine (Server, Manages images and containers), Docker CLI - Client (to interact with the docker Server, execute containers.. starting/stopping, deleting etc), GUI Client (same as CLI)

****Recall that docker allows us to package everything into an artifact which can be easily shared and moved (kind of like a zip folder, or jar file, etc)
- This can now be uploaded to a artifactory storage / downloaded on a server whenever needed
- We usually define everything we need inside of a Dockerfile
- Dockerfile Example:
```
FROM python:3.11-slim 
RUN pip install pandas
WORKDIR /appCOPY pipeline.py pipeline.pyCOPY wine_data.csv wine_data.csv
# Make port 80 available to the world outside this container
EXPOSE 80
ENTRYPOINT [ "python", "pipeline.py" ]
```