markhoerth opened a new issue, #12690:
URL: https://github.com/apache/gravitino/issues/12690
### What would you like to be improved?
### Why
Three changes going into 1.3.1 alter behavior for deployments already
running 1.3.0. Each of them looks like a healthy cluster after the
upgrade and fails later, which is the worst shape for an operator to
debug. They need to be stated in the release notes rather than
discovered.
### What needs to be in the notes
- [ ] **The OAuth credential is no longer written into the generated
CREATE CATALOG.** Multi-node deployments using oauth2 must configure
the environment-variable mapping on the coordinator and every worker
before upgrading. Without it, catalogs still register and metadata
still lists, so the cluster appears fine, and every query that
reaches a worker fails.
- [ ] **Iceberg REST routing is enabled by default in the Trino
connector.** A deployment on a released 1.3 server sees Iceberg
catalogs change behavior. The escape hatch is
`gravitino.iceberg.rest-routing-enabled`.
- [ ] **The Spark connector loses its silent fallback to hive/jdbc
translation** when no Iceberg REST endpoint is discovered, matching
the Trino behavior. Spark needs the same escape hatch, and both
belong in the notes as behavior changes.
### Related
- Spark connector items: #12687
### How should we improve?
_No response_
--
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]
