:microscope: nachet-backend 🌱

High level sequence diagram

sequenceDiagram

  title: High Level Sequence Diagram 1.0.0
  actor Client
  participant frontend
  participant backend
  participant EndpointAPI
  participant AzureStorageAPI

Client->>+frontend: getDirectoriesList()
frontend->>+backend: HTTP POST req.
backend->>+AzureStorageAPI: get_blobs()
AzureStorageAPI-->>-backend: blobListObject
backend-->>frontend: directories list res.
frontend-->>Client: display directories
Client->>frontend: handleInference()
frontend->>backend: HTTP POST req.
backend->>+AzureStorageAPI: upload_image(image)
AzureStorageAPI-->>-backend: imageBlobObject
backend->>+EndpointAPI: get_inference_result(image)
EndpointAPI-->>-backend: inference res.
backend->>backend: process inf. result
backend-->>frontend: inference res.
frontend-->>-Client: display inference res.
backend->>+AzureStorageAPI: (async) upload_inference_result(json)

Details

• The backend was built with the Quart framework
• Quart is an asyncio reimplementation of Flask
• All HTTP requests are handled in app.py in the root folder
• Azure Storage API calls are handled in the `azure_storage_api/azure_Storage_api.py
• Inference results from model endpoint are directly handled in model_inference/inference.py

RUNNING NACHET-BACKEND FROM DEVCONTAINER

When developping you first need to install the packages required.

This command must be run the first time you want to run the backend on your computer, but also every time you update the requirements.txt file and every time the datastore repo is updated

pip install -r requirements.txt

Then, you can run the backend while in the devcontainer by using this command:

hypercorn -b :8080 app:app

RUNNING NACHET-BACKEND AS A DOCKER CONTAINER

If you want to run the program as a Docker container (e.g., for production), use:

docker build -t nachet-backend .
docker run -p 8080:8080 -e PORT=8080 -v $(pwd):/app nachet-backend

RUNNING NACHET-BACKEND WITH THE FRONTEND IN DOCKER

If you want to run the frontend and backend together in Docker, use:

docker-compose up --build

You can then visit the web client at http://localhost:80. The backend will be build from the Dockerfile enabling preview of local changes and the frontend will be pulled from our Github registry.

TESTING NACHET-BACKEND

To test the program, use this command:

python -m unittest discover -s tests

ENVIRONMENT VARIABLES

Start by making a copy of .env.template and renaming it .env. For the backend to function, you will need to add the missing values:

• NACHET_AZURE_STORAGE_CONNECTION_STRING: Connection string to access external storage (Azure Blob Storage).
• NACHET_DATA: Url to access nachet-data repository
• NACHET_BLOB_PIPELINE_NAME: The name of the blob containing the pipeline.
• NACHET_BLOB_PIPELINE_VERSION: The version of the file containing the pipeline used.
• NACHET_BLOB_PIPELINE_DECRYPTION_KEY: The key to decrypt sensible data from the models.
• NACHET_VALID_EXTENSION: Contains the valid image extensions that are accepted by the backend
• NACHET_VALID_DIMENSION: Contains the valid dimensions for an image to be accepted in the backend.
• NACHET_MAX_CONTENT_LENGTH: Set the maximum size of the file that can be uploaded to the backend. Needs to be the same size as the client_max_body_size value set from the deployment in Howard.

DEPRECATED

• NACHET_MODEL_ENDPOINT_REST_URL: Endpoint to communicate with deployed model for inferencing.
• NACHET_MODEL_ENDPOINT_ACCESS_KEY: Key used when consuming online endpoint.
• NACHET_SUBSCRIPTION_ID: Was used to retrieve models metadata
• NACHET_WORKSPACE: Was used to retrieve models metadata
• NACHET_RESOURCE_GROUP: Was used to retrieve models metadata
• NACHET_MODEL: Was used to retrieve models metadata

DEPLOYING NACHET

If you need help deploying Nachet for your own needs, please contact us at cfia.ai-ia.acia@inspection.gc.ca.