gabrielmoris / docker

Docker core Concepts

Basics

Containers: Isolated "boxes" that contain our code and the environment needed to run it. 1 container must be focused in only 1 task. Images: Created by docker files. They contain the code and the envidonment. They are the blueprint of the container and are readonly

Basic commands

1. Build an Image docker build -t <NAME:TAG> <path of dockerfile>
2. Run a container based on a remote or local storage docker run --name <name> --rm -d <IMAGE> --rm => remove after stopping; -d detached mode
3. Push and pull to docker hub docker push <REPOSITORY/NAME:TAG> docker pull <REPOSITORY/NAME:TAG>

Bind Mounts & Volumes

Bind Mounts: Used to connect to host machine folders -v /local/path:/container/path

Volume: Used to persist the data -v NAME:/container/path

Networks and Container communication

Create a docker network: Using a network, containers can communicate with each other.

Docker-compose

Allows to pre-define the build and run steps in a .yaml file. It is specially helpful in multiple containers project.

To call this file I use in the .yaml path the command docker-compose up. To stop the services I use docker-compose down

Deploy considerations

• Replace Bind Mounts with Volumes or COPY
• Multiple containers might need multiple hosts
• Multi-stage Builds help with apps that need a build step

• Control : EC2 remote server, install Docker and run containers.
• Ease-of-use: ECS managed server, write the commands in AWS

Kubernetes

Is a open-source system for orchestrating container deployments. It helps with Automatic Deployment, Scaling & load balancing and Management

Core Concepts:

• Cluster: Set of Node machines which are running Containerized.
• Nodes: Physical or virtual machine which hosts one or multiple Pods and communicates with the Cluster
  • Master Node: Control Plane of the cluster. It manages the pods across Worker nodes
  • Worker Node: Machines Hosting Pods and running app containers
• Pods: Application Containers + needed Resources (for example Volumes)
• Containers: Regular Docker Containers
• Services: A Logical set of Pods with a unique Pod - / Continer- independent IP adress

Installation

1. Install chocolatey
2. in Docker Desktop enable kubernetes or run in CMD choco install kubernetes-cli
3. in CMD run choco install minikube
4. If I want to have a virtualbox (with docker I dont need) install virtualbox
5. To run the minikube in CMD run minikube start --driver=docker
6. Run minikube status to check if the minikube is runing or minikube delete to delete the virtual machine
7. Run minikube dashboard to open a tab with the cubernetes information in the browser

Kubernetes Objects

Can be declared imperatively or declaratively

Pod Object

Runs one or more containers and contain shared resources (volumes). They have a cluster-internal IP. They loose all the data when they are stopped (like containers)

Deployment Object

Controls one or multiple Pods