turborepo-s3-remote-cache

HTTP server to store Turborepo remote cache to an AWS S3 bucket.

⚠️ Do not use in production ⚠️

Configuration

Environment variable	Description
TURBOREPO_ALLOWED_TOKENS	Required. Allowed tokens for Turborepo authentication. Comma separated list
AWS_REGION	Required. Region of the S3 Bucket
AWS_ACCESS_KEY_ID	Required. Access key ID
AWS_SECRET_ACCESS_KEY	Required. Secret access key
AWS_ENDPOINT	Leave empty for default AWS endpoint. Customize for S3 compatible storage (e.g. min.io)
AWS_DISABLE_SSL	Disable SSL in the AWS SDK
AWS_S3_BUCKET	Default: turborepo-cache
AWS_S3_FORCE_PATH_STYLE	Use legacy path for S3 objects
GIN_TRUSTED_PROXIES	See Gin docs for more info. Comma separated list

For more info take a look at the AWS guide on configuring the SDK

Running

With Docker

docker run ghcr.io/acifani/turborepo-s3-remote-cache:latest

Check out all available tags in the Packages page.

From pre-built binaries

Download a binary from the Release page, unpackage it and run it.

Binaries are available for Windows, MacOS, Linux in amd64 and arm64

From source

git clone https://github.com/acifani/turborepo-s3-remote-cache.git
cd turborepo-s3-remote-cache
go build ./...
./turborepo-s3-remote-cache

Usage

You will need to configure the API endpoint and the auth token.

E.g.

export $TURBOREPO_TOKEN="some_t0k3n"
turbo run build --api="http://localhost:8080" --team="my-team" --token=$TURBOREPO_TOKEN

You will see this message if the remote cache has been correctly enabled.

• Remote computation caching enabled

If you provide a team name (recommended), the cache will be stored in a dedicated directory within the bucket.

Config file

You can also create a .turbo/config.json file and configure the API server and team id/slug there.

E.g.

{
  "apiUrl": "http://localhost:8080",
  "teamSlug": "my-team"
}

and then run

turbo run build --token=$TURBOREPO_TOKEN