Upload Garmin runs to Strava Lambda function

Architecture

Components:

• Lambda Function: performs the business logic
• Lambda Layer: provides the dependencies for the function
• CloudWatch Event: triggers the Lambda Function every minute
• DynamoDB: stores Strava credentials (cheapest option, ideally would use SecretsManager)
• CloudWatch Logs: stores Lambda Function logs
• IAM role/policy: allows the Lambda Function to write to CloudWatch Logs and read/write to DynamoDB

graph LR;
  trigger(CloudWatch Event);
  lambda("Lambda Function
  upload-strava-runs-to-garmin");
  lambda_layer[/"Lambda Layer
  python-dependencies-layer"/];
  db[("DynamoDB
  strava_tokens")];
  strava(Strava API);
  garmin(Garmin API);
  logs(CloudWatch Logs);
  trigger -- Trigger every minute --> lambda;
  lambda_layer --> lambda;
  lambda <-- Strava access tokens --> db;
  lambda <-- Fetch activities --> garmin;
  lambda <-- Upload activities --> strava;
  lambda --> logs;

Loading

Business logic

Note: Ephemeral storage is the /tmp directory in the Lambda function.

sequenceDiagram;
  participant lambda as Lambda Function;
  participant storage as Ephemeral storage;
  participant db as DynamoDB;
  participant strava as Strava API;
  participant garmin as Garmin API;
  lambda ->> storage: Check for Garmin credentials;
  alt Invalid credentials
    lambda ->> garmin: Login with credentials;
    lambda ->> storage: Store credentials;
  end
  lambda ->> db: Get Strava access token;
  alt Invalid credentials
    lambda ->> db: Get Strava refresh token;
    lambda ->> strava: Refresh access token;
    lambda ->> db: Store Strava tokens;
  end
  lambda ->> strava: Fetch Strava activities;
  lambda ->> garmin: Fetch Garmin activities;
  loop Garmin activities not in strava activities;
    lambda ->> garmin: Download FIT zip file;
    lambda ->> storage: Store FIT zip file;
    storage ->> storage: Extract zip contents;
    lambda ->> strava: Upload FIT file;
  end

Loading

Tools

• Terraform - installation
• PDM - installation
• Pre-commit - installation

Setup

You will need to set up an API application in Strava settings (see instructions). Use localhost:9321 as your path. Note your Client ID, Client Secret.

Once your API is set up, run

make get-token

and note your Refresh Token to be used for deployment.

You will also need to set up your AWS credentials to be used with Terraform. This can be done with the aws configure CLI (see instructions).

Deploying

The infrastructure is built and deployed with Terraform. Once you complete the setup, you will need to initialize Terraform. To do so, run

make tf-init

To see the changes that would be made to you account, run

make tf-plan

To publish the changes, run

make tf-apply

Development

To update code, first create an virtual environmnet with PDM by running

pdm venv create

And start it with

pdm venv activate

Install prod and dev dependencies with

pdm install

Before committing, make sure to run

pre-commit install

Linting

Formatting and linting use Ruff. Type checking is done with MyPy.

To format your code (both Terraform and Python), run

make format

To check your Python code with MyPy and Ruff, run

make check