Re: [I] [client] Expose all coordinator servers (leader and standbys) via the client API [fluss]

["MaheshS08 (via GitHub)" <gi...@apache.org>]

MaheshS08 commented on issue #4117:
URL: https://github.com/apache/fluss/issues/4117#issuecomment-5538068492

   @affo, I would like to work on this issue. I'll implement the feature to expose all coordinator servers (leader and standbys) via the client API with the following approach:
   
   1. Create a `CoordinatorRole` enum to distinguish between leader and standby coordinators
   2. Extend the proto messages (`PbCoordinatorServerInfo`) to include coordinator role and liveness information in `MetadataResponse`
   3. Modify the server-side logic to collect all registered coordinators from ZooKeeper (`/coordinators/ids`) with their current role (from `/coordinators/active`)
   4. Add a liveness indicator by checking coordinator presence in the election group
   5. Extend `Admin#getServerNodes()` to return all coordinators (leader + standbys) along with tablets servers
   6. Update client-side `Cluster` and `MetadataUpdater` to track and expose all coordinators
   
   This will enable operational tooling to:
   - Display the HA topology (how many coordinators, which is leader, standby status)
   - Verify standby coordinator liveness 
   - Support cluster readiness checks for standby coordinators (addressing #4105)
   
   The implementation will maintain backward compatibility while providing the new visibility into the coordinator HA setup.


-- 
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: issues-unsubscribe@fluss.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org