Docker Notes

What and Why of Docker ? Virtualization software Makes developing and deploying applications much easier Packages application with all the necessary dependencies, configuration, system tools and runtimes Portable artifact, easily shared and distributed Development before containers Development after containers Each developer needs to install and configure all services directly on their OS Installation process different for each OS environment Many steps where something can go wrong Own isolated environment Postgres packaged with all dependencies and configs Start service as a Docker container using 1 Docker command Command same for all services Standardized process for running any service on any local dev environment Easy to run different versions of the same app without conflicts Deployment before Containers Deployment after Containers Installations and configurations done directly on the server's OS Dependency version conflicts Human errors can happen Back and forth communications Instead of textual, everything is packaged inside the Docker artifact No configurations needed on the server (except Docker runtime) Less room for errors Install Docker runtime on the server Run Docker command to fetch and run the Docker artifacts Docker vs Virtual Machine An OS is made up of 2 layers i.e. OS Kernel and OS Application Layer Kernel is at the core of every OS Kernel interacts between hardware and software components Docker contains the OS application layer Services and apps are installed on top of that layer VMs contains both OS application layer and OS kernel layer Docker images are much smaller (Docker images are in MD while VM images in GB) Containers take seconds to start while VMs take minutes to start Docker is compatible only with Linux distros while VMs are compatible with all OS Docker was originally made for Linux However, Docker Desktop allows you to run Linux containers on Windows or MacOS Docker Desktop uses a Hypervisor layer with a lightweight Linux distro Install Docker Locally Got to the official Docker website Follow installation steps Docker Desktop includes Docker Engine A server with a long-running daemon process "Dockerd" Manages images and containers Docker CLI - Client CLI to interact with Docker Server Execute Docker commands to start/stop... containers GUI Client To manages your containers and images with a GUI Images vs Containers Images Docker Container An executable application artifact Includes app source code, but also complete environment configuration Add environment variables, create directories, files, etc. Immutable template that defines how a container will be realized You can run multiple containers from 1 image Actually starts the application A running instance of an image That's when the container environment is created Public and Private Registries Public Docker Registries Private Docker Registries Dockerhub largest public registry Anyone can search and download docker images You need to authenticate before accessing the registry All big cloud provider offer private registries: Amazon ECR, Google Container Registry, ect. Nexus (popular artifact repository manager) Docker Hub Registry vs Repository Docker Registry A service providing storage Collection of repositories Docker Repository Collection of related images with same name but different versions Docker Registries A storage and distribution system for Docker images Official images available form applications like Redis, Mongo,etc. Official images are maintained by the software authors or in collaboration with Docker community Docker hosts one of the biggest Docker Registry, called 'Docker Hub' Image Versioning Docker images are versioned Different versions are identified by tags Main Docker commands docker images #List all docker images docker ps #List all running containers #Pull an image from a registry docker pull {name}:{tag} docker pull nginx:1.23 #or docker pull nginx:latest To create a container with random name from given image(either local or in dockerhub) and starts it docker run {name}:{tag} docker run nginx:1.23 Run container in background and prints the container ID using -d or --detach docker run -d nginx:1.23 To view logs from service running inside the container (which are present at the time of execution) use : docker logs {container} docker logs 1e342f2 To stop a container docker stop {container} docker stop 1e342f2 Run Containers docker run command Creates a new container Doesn't re-use previous container use -a pr --all to List all container (stopped and running) docker ps -a to stop a container : docker stop {container} to give name to a container when creating a new

May 2, 2025 - 14:05

What and Why of Docker ?

Virtualization software
Makes developing and deploying applications much easier
Packages application with all the necessary dependencies, configuration, system tools and runtimes
Portable artifact, easily shared and distributed

Development before containers	Development after containers
Each developer needs to install and configure all services directly on their OS Installation process different for each OS environment Many steps where something can go wrong	Own isolated environment Postgres packaged with all dependencies and configs Start service as a Docker container using 1 Docker command Command same for all services Standardized process for running any service on any local dev environment Easy to run different versions of the same app without conflicts

Deployment before Containers	Deployment after Containers
Installations and configurations done directly on the server's OS Dependency version conflicts Human errors can happen Back and forth communications	Instead of textual, everything is packaged inside the Docker artifact No configurations needed on the server (except Docker runtime) Less room for errors Install Docker runtime on the server Run Docker command to fetch and run the Docker artifacts

Docker vs Virtual Machine

An OS is made up of 2 layers i.e. OS Kernel and OS Application Layer
Kernel is at the core of every OS
Kernel interacts between hardware and software components
Docker contains the OS application layer
Services and apps are installed on top of that layer
VMs contains both OS application layer and OS kernel layer
Docker images are much smaller (Docker images are in MD while VM images in GB)
Containers take seconds to start while VMs take minutes to start
Docker is compatible only with Linux distros while VMs are compatible with all OS
Docker was originally made for Linux
However, Docker Desktop allows you to run Linux containers on Windows or MacOS
Docker Desktop uses a Hypervisor layer with a lightweight Linux distro

Install Docker Locally

Got to the official Docker website
Follow installation steps

Docker Desktop includes

Docker Engine
- A server with a long-running daemon process "Dockerd"
- Manages images and containers
Docker CLI - Client
- CLI to interact with Docker Server
- Execute Docker commands to start/stop... containers
GUI Client
- To manages your containers and images with a GUI

Images vs Containers

Images	Docker Container
An executable application artifact Includes app source code, but also complete environment configuration Add environment variables, create directories, files, etc. Immutable template that defines how a container will be realized You can run multiple containers from 1 image	Actually starts the application A running instance of an image That's when the container environment is created

Public and Private Registries

Public Docker Registries	Private Docker Registries
Dockerhub largest public registry Anyone can search and download docker images	You need to authenticate before accessing the registry All big cloud provider offer private registries: Amazon ECR, Google Container Registry, ect. Nexus (popular artifact repository manager) Docker Hub

Registry vs Repository

Docker Registry

A service providing storage
Collection of repositories

Docker Repository

Collection of related images with same name but different versions

Docker Registries

A storage and distribution system for Docker images
Official images available form applications like Redis, Mongo,etc.
Official images are maintained by the software authors or in collaboration with Docker community
Docker hosts one of the biggest Docker Registry, called 'Docker Hub'

Image Versioning

Docker images are versioned
Different versions are identified by tags

Main Docker commands

docker images #List all docker images

docker ps #List all running containers

#Pull an image from a registry
docker pull {name}:{tag}

docker pull nginx:1.23
#or
docker pull nginx:latest

To create a container with random name from given image(either local or in dockerhub) and starts it

docker run {name}:{tag}

docker run nginx:1.23

Run container in background and prints the container ID using -d or --detach

docker run -d nginx:1.23

To view logs from service running inside the container (which are present at the time of execution) use :

docker logs {container}
docker logs 1e342f2

To stop a container

docker stop {container}
docker stop 1e342f2

Run Containers

docker run command Creates a new container
Doesn't re-use previous container
use -a pr --all to List all container (stopped and running)

docker ps -a

to stop a container :

docker stop {container}

to give name to a container when creating a new container, use:

docker run --name {name} {name}:{tag}

docker run --name web-app nginx:1.23

to stare one or more stopped containers, use:

docker start {container}

Port Binding

Application inside container runs in an isolated Docker network
This allows us to run the same app running on the same port multiple times
We need to expose the container port to the host (the machine the container runs on)
use -p or --publish to publish a container's port to the host

docker run -d -p {HOST_PORT}:{CONTAINER_PORT} {name}:{tag}

docker run -d -p 9000:80 nginx:1.23

Create own Image (Dockerfile)

Dockerfile is text document that contains commands to assemble an image
Docker can then build an image by reading those instruction

Structure of Dockerfile

Dockerfile start from a parent image or base image
You choose the base image, depending on which tool you need to have available, for example node, tomcat, python,etc.
Dockerfile must begin with a FROM instruction
Build this image from the specified image
RUN command will execute any command in a shell inside the container environment
COPY command copies files or directories from and adds them to the filesystem of the container at the path

COPY {src} {dest}

While RUN is executed on container, COPY is execute on host
WORKDIR sets the working directory for all following commands
Like changing into a directory using cd
CMD command is for the instruction that is to be execute when a Docker container starts
There can only be one CMD command in a dockerfile.

Example

Let our node.js project file structure be

.
├── src/
│   ├── server.js
│   └── Dockerfile
└── package.json

Then our Dockerfile will be :

FROM node:19-alpine

COPY package.json /app/
COPY src /app/

WORKDIR /app

RUN npm install

CMD ["node","server.js"]

Building the image

To build the image use the following command:

docker build -t {name}

docker build -t node-app

The -t or --tag is used to set a name and optionally a tag in the "name:tag" format