This example project serves simple SVG labels using Lambda, such as the one below:
See repolabels.net for a live version of this example, and generate labels for your own repositories.
This example demonstrates how to:
- Bind method parameters to URL path components (requests are processed as
{owner}/{repo}/{template}) - Customise response content types (this serves
image/svg+xml) - Manage development/testing/production API versions
- Connect to a third party REST API (github)
- store keys for development/testing/production access
- Protect API access using API keys
- reduce costs by caching results using CloudFront
Check out the 20 minute video tutorial How to make a cheap, scalable image server with AWS for a detailed walk-through.
Create the development version of the API using npm run start, and deploy the production version using npm run release.
Github limits the rate of unauthenticated requests severely, this example allows you to optionally specify two stage variables to authenticate requests to GitHub, and increase rate limits. Check out your deployed API in the AWS web console, and add two stage variables to the production stage - githubClientId and githubSecret. Create a GitHub application to get access keys. You can optionally also create a development GitHub application and store its keys in the development stage.
You can change the development version and run npm run deploy, and you will see that the development version changed, but the production version did not. For example, uncomment apiKeyRequired: true in the last line of web.js before deploying. This will cause the API calls to require an API key before allowing the request. See Use an API Key in API Gateway for more information on how to create an use an API key.
To push the currently released version to production, run npm run release again.