Skip to content

Latest commit

 

History

History
112 lines (83 loc) · 3.77 KB

DEPLOYMENT.md

File metadata and controls

112 lines (83 loc) · 3.77 KB

Deployment on AWS lambda

This document describes how to deploy the API on AWS Lambda. If you do not want to use AWS S3, you can skip the final section about linking the API to AWS S3.

Setting up the API on your AWS account

Prerequisites:

API (AWS Lambda)

Creating a Lambda function on your AWS account

Create a Lambda function on your AWS account to run the API. You can use the console to create your lambda function. This will automatically link a AWSLambdaBasicExecutionRole to it.

After creation, alter the following for your lambda function:

  • Set Runtime to Python 3.10 (Code > Runtime Settings)
  • Set the Runtime Handler to app.main.handler (Code > Runtime Settings)
  • Enable function URL with Auth type NONE (Configuration > Function URL)
    • Explanation: authentication will be done by FastAPI

Configure environment variables for your lambda function

On the lambda console of your function, head over to 'Configuration -> Environment variables' and add the following:

  • set API_KEY_READ_ACCESS to <a_random_string>
  • set API_KEY_WRITE_ACCESS to <a_different_random_string>
  • set USE_LAMBDA_DEV_SETTINGS to true

Note: The LAMBDA_DEV_SETTINGS are not very restrictive on security. You should disable this when deploying for production.

Install the dependencies on your local machine using poetry

Create a python 3.10 environment:

poetry env use 3.10

Then install the dependencies:

poetry install

Since in-project = true is configured in poetry.toml, this will create a virtual environment within the project folder.

Build the lambda package

From the root of the project, run:

./lambda_zip.sh

This script will zip the whole project into package.zip in the root of the project.

Upload the lambda package to your lambda function

From the root of the project, run:

./lambda_deploy.sh
  • Note: This script assumes you have a lambda function on AWS with the same name as the current directory.

Alternatively, you can use the AWS console to upload the zip file.

Done

You can now test your API on AWS Lambda. The URL of your API is available on the lambda console under 'Configuration > Function URL'.

Visit https://<your_function_url>/docs to see the Swagger page of your API.

Linking your API to AWS S3

Bucket creation

Create a S3 Bucket on your AWS account to store the json files in. Either use the console or AWS CLI:

aws s3 mb s3://<your_unique_bucket_name>

Extra environment variables for your lambda function

On the lambda console of your function, head over to 'Configuration -> Environment variables' and add the following:

  • set S3_BUCKET_NAME to <your_unique_bucket_name>
  • set LOCAL_ASSETS_DIR to /tmp

Extending the linked IAM role with S3 permissions

Next, head to 'Configuration > Permissions' and click the link to the IAM role.

Attach the following S3 policy to the role:

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Effect": "Allow",
            "Action": "s3:*",
            "Resource": [
                "arn:aws:s3:::<your_unique_bucket_name>/assets",
                "arn:aws:s3:::<your_unique_bucket_name>/assets/*"
            ]
        }
    ]
}

Done

Your API now has read/write access to your S3 bucket. You can now test your API on AWS Lambda. The URL of your API is available on the lambda console under 'Configuration > Function URL'.

Visit https://<your_function_url>/docs to see the Swagger page of your API.