wuchong commented on issue #2683:
URL: https://github.com/apache/fluss/issues/2683#issuecomment-3906361501

   @MehulBatra I think you raised a good question. I recall we refactored the 
quickstart image into a standard Flink image plus manual JAR downloads 
primarily to avoid combinatorial explosion of Docker images for different 
lakehouse formats.
   
   I think one practical approach could be:
   
   - Keep Fluss-related dependencies pre-placed in `flink/lib` so that the 
“Quickstart with Flink” scenario works out of the box—no Docker customization 
or extra steps needed.
   - Place Paimon and Iceberg dependencies in separate directories like 
`flink/paimon/` and `flink/iceberg/`.
   - For “Quickstart with Paimon Lakehouse,” move the contents of 
`flink/paimon/` into `flink/lib/` before starting Flink. This can be achieved 
via a custom `command` or `entrypoint` in the `docker-compose.yml`.
   - If the initialization is complex, we can pre-bundle an initialization 
script (e.g., `init_paimon.sh`) inside the image. Users would simply invoke it 
before launching the main process—for example:  
     ```yaml
     command: ["/opt/flink/init_paimon.sh", "&&", "bin/jobmanager.sh"]
     ```
   
   This way, we preserve the simplicity of the default Fluss quickstart while 
still supporting optional lakehouse integrations through lightweight, 
script-driven setup—without requiring users to build or manage multiple Docker 
images.


-- 
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]

Reply via email to