[I] [flink] Implement Fluss full-cache lookup join provider [fluss]

["loserwang1024 (via GitHub)" <gi...@apache.org>]

loserwang1024 opened a new issue, #3726:
URL: https://github.com/apache/fluss/issues/3726

   ### Search before asking
   
   - [x] I searched in the [issues](https://github.com/apache/fluss/issues) and found nothing similar.
   
   
   ### Motivation
   
   Currently, Fluss Flink Connector supports lookup join through per-record lookup RPCs, optionally with Flink's partial lookup cache. In this mode, the connector does not preload dimension table data. Each cache miss triggers a lookup request to Fluss TabletServer.
   
   This works for small traffic or low-cardinality lookup workloads, but it can become expensive in production when lookup QPS is high or the dimension table is relatively stable.
   
   This issue aims to implement a Fluss full-cache lookup join provider, so lookup data can be loaded into local cache before serving lookup requests.
   
   For many dimension table scenarios, users prefer full-cache lookup join:
   
   ```sql
   SELECT s.order_id, s.user_id, d.user_name, d.level
   FROM order_stream AS s
   LEFT JOIN user_dim FOR SYSTEM_TIME AS OF s.proc_time AS d
   ON s.user_id = d.user_id;
   ```
   
   If user_dim is fully cached locally, lookup requests can be served without per-record RPC to Fluss, reducing latency and TabletServer pressure.
   
   
   ### Solution
   
   _No response_
   
   ### Anything else?
   
   _No response_
   
   ### Willingness to contribute
   
   - [ ] I'm willing to submit a PR!


-- 
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.apache.org

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