# Docker & Kubernetes introduction (CARTA)
- Docker
-- Container key features
-- Container versus Virtual Machine(VM)
-- Basic command line
-- CARTA CI/CD
- Kubernetes (k8S)
-- Official definition
-- What problems does k8S solve?
-- What k8S offer?
-- Basic four components

## Docker
Docker is a set of platform as a service (PaaS) products that use **OS-level virtualization** (e.g. Linux kernel namespace) to deliver software in packages called **containers**.

Docker is a tool that is used to automate the deployment of applications in lightweight containers so that applications can work efficiently in different environments in isolation.

#### Container key features
- Consider as individual OS-level virtualization
- Containers sharing the same Linux kernel
- Each Container is isolated process
- light-weight, can run several containers simultaneously.
- Easy to deploy

#### Container versus Virtual Machine(VM)

<span style="background-color: #FFFF00">
**MacOS Docker can have Ubuntu Linux kernel, however Linux kernel does not have MacOSX kernel. This is the main reason that our MacOS CARTA CI/CD are all native (MacOS11: acdcs-Mac-Pro; MacOS12: alm1ac; MacOS13: almac)**
</span>

**Official Docker image (e.g. ubuntu and centos) can be found in the [docker hub](https://hub.docker.com/)**

#### Basic command line
1. We must have **Dockerfile**, we have Dockerfile in our CARTA backend repository ([link](https://github.com/CARTAvis/carta-backend/tree/dev/Dockerfiles))
2. Using the Dockerfile to build the **image**

## For example: I am building Ubuntu20.04 from Dockerfile and giving a tag: mylin/carta-ubunutu20.04:v1
docker build -f Dockerfile-ubuntu-20.04 . --tag mylin/carta-ubunutu20.04:v1

3. Check the image list

docker images ls


4. Using the image to build the container

docker run -itd -v $PWD:/mount -p 3002:3002 --name carta-ubunutu20.04 mylin/carta-ubunutu20.04:v1

6. Check the container list

docker ps -a


7. Access the container

docker exec -it [CONTAINER ID, e.g. 2d64a179fe94] bash

#### CARTA CI/CD
We **used** Jankins blue-ocean plugin to run the carta CI/CD pipeline on the server (Linux in the containers and MacOS are native). (Now is removed from our project)

Now CI/CD pipelines are using **Apptainer (formerly Signularity)** with github action from Dec 2023.
It is very similar to Docker but the command lines are slightly different.

* Dockerfile <-> ubuntu-2004-dec2023.def (under cartaviewer server `/home/acdc/apptainer`)
* Container <-> ubuntu-2004-dec2023.sif

Using the *.def to build the latest version of *.sif

apptainer build testing.sif ubuntu-2004-dec2023.def

* [Docker]`docker image ls` <-> [Signularity]`ls *.sif`
* [Docker]`docker ps -a` <-> [Signularity]`singularity instance list`
* [Docker]`docker exec -it [CONTAINER ID] bash` <-> [Signularity]`apptainer shell instance://[INSTANCE NAME]`

## Kubernetes
Motivation: Why Heng-tai implemented k8s in the carta-backend ICD tests
([link](https://github.com/CARTAvis/carta-backend-ICD-test/tree/master/k8s))

[Youtube k8S Reference](https://www.youtube.com/watch?v=s_o8dwzRlu4)
[Game article Reference](https://medium.com/my-games-company/no-need-to-pause-how-we-update-games-without-downtime-7ebcff1ce813)

#### Official definition
* Open source container orchestration framework(tool)
* Developed by Google
* Managing containerized applications in different deployment environments(e.g. physical machine, virtual machine, cloud, etc.)

#### What problems does K8S solve?
* Services (contains serveral applications) from Monolith to microservices (small and independent)
* Increased usage of containers
* A proper way to manage those containers

#### What k8S offer?
* High avaliability (aka NO downtime), for example: the game services (Love and Deepspace from China 2024) offer the version update without downtime. Or Cisco offers high avaliability services ([link](https://www.cisco.com/c/en/us/solutions/hybrid-work/what-is-high-availability.html))
* Scalability or high performance (many users means high pressure data load)
* Disaster recovery: backup and restore (e.g. self healing)

#### Basic four components
([official website](https://kubernetes.io/docs/concepts/overview/components/))
* cluster
> When you deploy Kubernetes, you get a cluster.
* node(s)
> A Kubernetes cluster consists of a set of worker machines, called nodes, that run containerized applications. Every cluster has at least one worker node.
* pod(s)
> The worker node(s) host the Pods that are the components of the application workload.
* control plane (aka master node)
> The control plane manages the worker nodes and the Pods in the cluster. In production environments, the control plane usually runs across multiple computers and a cluster usually runs multiple nodes, providing fault-tolerance and high availability.