turbo-repo-remote-cache-rs

Fast turbo remote cache server written in Rust.

if you are using turbo and you want to have a self-hosted remote cache server this is for you.

1. It is fast ⚡️.
2. It supports multiple object storage providers (S3, GCP, Azure, and file).

Installation

• Using Cargo (Linux/macOS/Windows)

cargo install turbo-remote-cache-rs

• Using Npm (Linux/macOS/Windows)

npm install -g turbo-remote-cache-rs

• Download the latest binary from release page

Usage

1. Copy .env.example to same directory as the binary and rename it to .env.
2. Edit .env file to your needs (TURBO_TOKENS, BUCKET_NAME are required) please refer to .env.example for more information about required vars for each cloud storage provider.
3. Run the binary.

  turbo-repo-remote-cache-rs

4. Provide the url to turbo cli

turbo build --api="http://localhost:4000" --token=<token> --team=<team>

you can also set teamId and apiUrl .turbo/config.json in the root of your project.

5. Enjoy your self-hosted remote cache and faster builds 🚀.

Docker

1. Create a docker file.

FROM salamaashoush/turbo-remote-cache-rs
ENV PORT=4000
ENV FS_PATH=/tmp
ENV BUCKET_NAME=cache
ENV TURBO_TOKENS="token1,token2,token3"
EXPOSE 4000
CMD ["turbo-remote-cache-rs"]

2. Build the image.

docker build -t your-image .

3. Run the image.

docker run -p 4000:4000 -v ./cache:/tmp/cache your-image

4. Provide the URL to turbo cli

turbo build --api="http://localhost:4000" --token=<token> --team=<team>

you can also set teamId and apiUrl .turbo/config.json in the root of your project.

Kubernetes

See example in examples/k8s, Don't forget to change the spec and env vars for your needs before applying it (NOTE that it is just an example and it is not production ready).

Environment Variables

Required

Name	Description	Default
TURBO_TOKENS	Comma separated list of turbo tokens that are allowed to access the cache.	""
BUCKET_NAME	Name of the bucket to store the cache in.	"cache"
STORAGE_PROVIDER	Storage provider to use. s3, azure, gcs, file or memory	"memory"

File Storage Provider

Name	Description	Default
FS_PATH	Path to store the cache in.	os temp dir

S3 Storage Provider

Name	Description	Default
AWS_ACCESS_KEY_ID	AWS access key id.	""
AWS_SECRET_ACCESS_KEY	AWS secret access key.	""
AWS_DEFAULT_REGION	AWS region.	""
AWS_ENDPOINT	AWS endpoint.	""
AWS_SESSION_TOKEN	AWS session token	""
AWS_CONTAINER_CREDENTIALS_RELATIVE_URI	AWS bucket endpoint.	""
AWS_ALLOW_HTTP	set to “true” to permit HTTP connections without TLS.	false
AWS_PROFILE	set profile name, requires aws_profile feature enabled	""

Azure Storage Provider

Name	Description	Default
AZURE_STORAGE_ACCOUNT_NAME	storage account name.	""
AZURE_STORAGE_ACCOUNT_KEY	storage account master key	""
AZURE_STORAGE_ACCESS_KEY	alias for AZURE_STORAGE_ACCOUNT_KEY	""
AZURE_STORAGE_CLIENT_ID	client id for service principal authorization	""
AZURE_STORAGE_CLIENT_SECRET	client secret for service principal authorization	""
AZURE_STORAGE_TENANT_ID	tenant id used in oauth flows	""

GCS Storage Provider

Name	Description	Default
GOOGLE_SERVICE_ACCOUNT	location of service account file	""
GOOGLE_SERVICE_ACCOUNT_PATH	(alias) location of service account file	""
SERVICE_ACCOUNT	(alias) location of service account file	""
GOOGLE_SERVICE_ACCOUNT_KEY	JSON serialized service account key	""
GOOGLE_BUCKET	bucket name	""
GOOGLE_BUCKET_NAME	(alias) bucket name	""

Todo

• Support turbo headers x-artifact-duration, x-artifact-tag, x-artifact-client-ci and x-artifact-client-interactive right now those are ignored and they don't affect the cache.
• Add nx cloud support.
• Add more advanced authentication support.
• Maybe having a dashboard to manage teams and projects would be nice.