Simple k8S Demo

Check the deployed demo demo

SETUP

STEP 1. CREATE .env FILE

  # Postgres
  DB_ENGINE=django.db.backends.postgresql
  DB_HOST=***
  DB_PORT=5432
  DB_NAME=***
  DB_USER=***
  DB_PASSWORD=***
  

  # Azure
  AZURE_ACCOUNT_NAME=***
  AZURE_ACCOUNT_KEY=***
  AZURE_CONTAINER_NAME=***
  AZURE_STORAGE_CONNECTION_STRING=***
  AZURE_QUEUES_NAME=

Note: put the .env in root dir

STEP 2: RUN WITH DOCKER COMPOSE LOCALLY

sudo rm -rf db/
docker-compose build
docker-compose up

ENDPOINT DETAIL

IMAGE UPLOAD ENDPOINT

endpoint : /apis/upload/ method: POST

request

• Headers

  Content-Type: multipart/form-data

• Parameters

  image: The image file to be uploaded. This should be a file input.

sample response

{
    "id": 37,
    "original_image_url": "https://task.blob.core.windows.net/images/original/eccd5e89-ef9e-49bf-a6fc-78c9374b179d.jpeg",
    "processed_image_url": null,
    "status": "pending",
    "upload_time": "2024-06-10T15:13:51.837204Z",
    "processing_start_time": null,
    "processing_end_time": null
}

ALL IMAGE TASK DETAILS

endpoint : /apis/tasks/

method: GET

sample response

[
    {
        "id": 1,
        "original_image_url": "https://task.blob.core.windows.net/images/original/58575466-37a8-4807-aba3-3a33ea8a21bb.jpg",
        "processed_image_url": "https://task.blob.core.windows.net/images/processed/58575466-37a8-4807-aba3-3a33ea8a21bb_processed.jpg",
        "status": "completed",
        "upload_time": "2024-06-10T15:36:30.730933Z",
        "processing_start_time": "2024-06-10T15:36:35.579294Z",
        "processing_end_time": "2024-06-10T15:36:35.725134Z"
    },
    {
        "id": 2,
        "original_image_url": "https://task.blob.core.windows.net/images/original/997036b3-65c1-400b-ac19-003144ca65fe.jpg",
        "processed_image_url": "https://task.blob.core.windows.net/images/processed/997036b3-65c1-400b-ac19-003144ca65fe_processed.jpg",
        "status": "completed",
        "upload_time": "2024-06-10T15:36:45.348888Z",
        "processing_start_time": "2024-06-10T15:36:50.560316Z",
        "processing_end_time": "2024-06-10T15:36:50.958846Z"
    }
]

SPECIFIC IMAGE TASK DETAILS

endpoint : /apis/tasks/<task_id:int>

method: GET

sample response

{
    "id": 1,
    "original_image_url": "https://task.blob.core.windows.net/images/original/58575466-37a8-4807-aba3-3a33ea8a21bb.jpg",
    "processed_image_url": "https://task.blob.core.windows.net/images/processed/58575466-37a8-4807-aba3-3a33ea8a21bb_processed.jpg",
    "status": "completed",
    "upload_time": "2024-06-10T15:36:30.730933Z",
    "processing_start_time": "2024-06-10T15:36:35.579294Z",
    "processing_end_time": "2024-06-10T15:36:35.725134Z"
}

CI/CD PIPELINE

Pipeline Overview The pipeline is triggered by events:

• A push to the main branch.

Permissions

The pipeline requires write permissions to the repository contents.

Jobs

The pipeline consists of the following jobs:

1. Build

2. Test

3. Security

4. Merge

   • Runs on: ubuntu-latest

   • Needs: build, test, security

   • Steps:

     • Checks out the code from the repository.
     • Configures Git with GitHub Actions credentials.
     • Merges the code from the pull request to the main branch.
     • Authenticates with GitHub and pushes the merged code to the main branch.

5. Docker Push

   • Runs on: ubuntu-latest
   • Needs: merge
   • Steps:
     • Checks out the code from the repository.
     • Builds a Docker image for the Django application.
     • Logs in to Docker Hub using the provided credentials.
     • Pushes the Docker image to Docker Hub.

6. Deploy

   • Runs on: ubuntu-latest

   • Needs: docker-push

   • Steps:

     • Azure CLI login
     • Setup Kubectl

GITHUB ACTION SECRET

• GH_TOKEN: GitHub token for authentication.
• DOCKER_USERNAME: Docker Hub username.
• DOCKER_PASSWORD: Docker Hub password.
• AZURE_CREDENTIAL: Azure cli authentication.
• AZURE_RESOURCE_GROUP: Azure resource group name.
• AZURE_AKS_CLUSTER_NAME: Azure AKS cluster name

MANUAL DEPLOYMENT

az login 

az aks get-credentials --resource-group AI --name backend-deployment-cluster

docker build -t DOCKER_USERNAME/django-app:latest .
docker push DOCKER_USERNAME/django-app:latest

kubectl apply -f k8s/backend_service/.
kubectl apply -f k8s/database_service/.