Configure S3 Access
Configure credentials and run an installed plugin against an S3 or S3-compatible product target.
Prerequisites
- Firecube and a compatible plugin installed in the same environment.
- The Firecube environment activated.
- Source data accepted by the plugin.
- An S3 bucket and credentials supplied by the storage provider.
Configure The Current Shell
Replace the values below. Omit the endpoint when the provider uses its standard S3 endpoint. Set the region only when the provider requires one.
export FIRECUBE_ENDPOINT_URL="https://your-s3-endpoint.example.com"
export FIRECUBE_REGION="your-region"
export FIRECUBE_ACCESS_KEY="your-access-key"
export FIRECUBE_SECRET_KEY="your-secret-key"
Environment variables have higher precedence than config.toml. Explicit CLI
values still win over environment variables.
Store Persistent Settings
Firecube reads ~/.config/firecube/config.toml by default. Keep the file
private when it contains credentials.
[storage]
endpoint_url = "https://your-s3-endpoint.example.com"
region = "your-region"
access_key = "your-access-key"
secret_key = "your-secret-key"
Run An S3 Ingestion
Replace PLUGIN_NAME, the source path, bucket, and product name with values
supported by the installed plugin:
firecube ingest PLUGIN_NAME \
--input-data /absolute/path/to/source-data \
--target s3://your-bucket/products/my_product.zarr \
--product-name my_product \
--storage-type s3 \
--storage-driver fsspec \
--output-format zarr \
--write-mode staged
The final summary should report
s3://your-bucket/products/my_product.zarr as the stored_at value.
Verify The Product
Inspect the product's control-plane records:
See the Configuration Reference for all storage fields and the Configuration Model for precedence rules.
Next Steps
- Create an Intake Catalog: publish catalog entries for an existing Zarr or Parquet product.
- Storage Drivers: compare supported S3 driver behavior.
- Recover Runs And Claims: recover after a failed process once no writer is still active.