diqiu50 opened a new issue, #12554: URL: https://github.com/apache/gravitino/issues/12554
### What would you like to be improved? `credential-providers=s3-token` does not work for a `lakehouse-iceberg` catalog loaded through the Gravitino Trino connector. No credential is obtained and S3 access fails. Two causes: the connector fetches credentials once at catalog registration with no table path, while `s3-token` requires one; and `catalog-backend=jdbc` is translated into Trino's `iceberg.catalog.type=jdbc`, which has nowhere to put an STS session token. Trino only consumes vended credentials in its `rest` catalog type. ### How should we improve? Load `lakehouse-iceberg` catalogs through the Iceberg REST server instead. `catalog-backend` describes how Gravitino stores metadata; it should not decide the engine's access path. The IRC endpoint and its authentication are deployment topology, so they belong in the Trino connector config, set once per cluster. Deployments without an IRC need a switch to fall back. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
