Skip to content

Latest commit

 

History

History
53 lines (42 loc) · 3.29 KB

README.md

File metadata and controls

53 lines (42 loc) · 3.29 KB
Raw

Deploying a Flask API and MySQL server on Kubernetes

This repo contains code that

  1. Deploys a MySQL server on a Kubernetes cluster
  2. Attaches a persistent volume to it, so the data remains contained if pods are restarting
  3. Deploys a Flask API to add, delete and modify users in the MySQL database

Prerequisites

  1. Have Docker and the Kubernetes CLI (kubectl) installed together with Minikube (https://kubernetes.io/docs/tasks/tools/)

Getting started

  1. Clone the repository
  2. Configure Docker to use the Docker daemon in your kubernetes cluster via your terminal: eval $(minikube docker-env)
  3. Pull the latest mysql image from Dockerhub: Docker pull mysql
  4. Build a kubernetes-api image with the Dockerfile in this repo: Docker build . -t flask-api

Secrets

Kubernetes Secrets can store and manage sensitive information. For this example we will define a password for the root user of the MySQL server using the Opaque secret type. For more info: https://kubernetes.io/docs/concepts/configuration/secret/

  1. Encode your password in your terminal: echo -n super-secret-passwod | base64
  2. Add the output to the flakapi-secrets.yml file at the db_root_password field

Deployments

Get the secrets, persistent volume in place and apply the deployments for the MySQL database and Flask API

  1. Add the secrets to your kubernetes cluster: kubectl apply -f flaskapi-secrets.yml
  2. Create the persistent volume and persistent volume claim for the database: kubectl apply -f mysql-pv.yml
  3. Create the MySQL deployment: kubectl apply -f mysql-deployment.yml
  4. Create the Flask API deployment: kubectl apply -f flaskapp-deployment.yml

You can check the status of the pods, services and deployments.

Creating database and schema

The API can only be used if the proper database and schemas are set. This can be done via the terminal.

  1. Connect to your MySQL database by setting up a temporary pod as a mysql-client: kubectl run -it --rm --image=mysql --restart=Never mysql-client -- mysql --host mysql --password=<super-secret-password> make sure to enter the (decoded) password specified in the flaskapi-secrets.yml
  2. Create the database and table
    1. CREATE DATABASE flaskapi;
    2. USE flaskapi;
    3. CREATE TABLE users(user_id INT PRIMARY KEY AUTO_INCREMENT, user_name VARCHAR(255), user_email VARCHAR(255), user_password VARCHAR(255));

Expose the API

The API can be accessed by exposing it using minikube: minikube service flask-service. This will return a URL. If you paste this to your browser you will see the hello world message. You can use this service_URL to make requests to the API

Start making requests

Now you can use the API to CRUD your database

  1. add a user: curl -H "Content-Type: application/json" -d '{"name": "<user_name>", "email": "<user_email>", "pwd": "<user_password>"}' <service_URL>/create
  2. get all users: curl <service_URL>/users
  3. get information of a specific user: curl <service_URL>/user/<user_id>
  4. delete a user by user_id: curl -H "Content-Type: application/json" <service_URL>/delete/<user_id>
  5. update a user's information: curl -H "Content-Type: application/json" -d {"name": "<user_name>", "email": "<user_email>", "pwd": "<user_password>", "user_id": <user_id>} <service_URL>/update