Serverless Puppeteer using AWS Lambda Layers

<p> <img src="https://img.shields.io/badge/version-1.1.0-blue.svg?cacheSeconds=2592000" /> <a href="https://twitter.com/rafalwilinski"> <img alt="Twitter: rafalwilinski" src="https://img.shields.io/twitter/follow/rafalwilinski.svg?style=social" target="_blank" /> </a> </p>

Recently AWS introduced Layers which enables sharing common code between functions and working with large dependencies (such as headless chrome) much easier.

This project example returns a screenshot of page requested via ?address= query parameter.

Usage

It is very important to tell NPM to skip installing chromium from puppeteer package. To do so, installing dependencies should be done using this command:

shCopy
PUPPETEER_SKIP_CHROMIUM_DOWNLOAD=1 npm install

or

shCopy
PUPPETEER_SKIP_CHROMIUM_DOWNLOAD=1 yarn

Without this, your function size would be more than 200MB which is way too much.

After that, you can simply deploy your funciton.

shCopy
sls deploy

curl <DEPLOYED_ENDPOINT>?address=https://google.com

or apply this ARN to your function directly: arn:aws:lambda:us-east-1:085108115628:layer:chrome:6

Bundling latest Chromium on your own

Prerequisites:

• brotli

shCopy
./download-latest.sh

Credits for this solution go to San Kumar

Author

👤 Rafal Wilinski

• Twitter: @rafalwilinski
• Github: @RafalWilinski

Show your support

Give a ⭐️ if this project helped you!