zhangyue19921010 opened a new issue, #11281:
URL: https://github.com/apache/paimon/issues/11281

   ### Search before asking
   
   - [x] I searched in the issues and found no report of ignored filesystem 
rename results.
   
   ### Paimon version
   
   Current master, `3a8e01f35d8fc1391f86286001a98340d65d7f80` (2.2-SNAPSHOT).
   
   ### Compute Engine
   
   Java table API. Flink and Spark rename procedures delegate to the same core 
methods.
   
   ### Minimal reproduce step
   
   1. Create a filesystem-backed table, commit a row, and create a tag or a 
branch named `old`.
   2. Have `FileIO.rename(source, target)` return `false` when renaming it to 
`new`. This is a supported failure result; `LocalFileIO` uses it for 
access-denied errors and an existing destination, including a destination 
created after the preliminary check.
   3. Call `table.renameTag("old", "new")` or `table.renameBranch("old", 
"new")`.
   
   The persisted-table regression injects that result after normal tag/branch 
creation. Both calls return normally even though `old` remains and `new` was 
not created.
   
   ### What doesn't meet your expectations?
   
   The operation should report the failed rename. Both managers ignore the 
boolean return value, so the Flink/Spark procedures can report success for a 
rename that did not happen. Check the result and fail with the source and 
destination names.
   
   ### Anything else?
   
   Existing IOException handling and successful renames can remain unchanged. 
Catalog-managed branch renames use a separate path.
   
   ### Are you willing to submit a PR?
   
   - [x] 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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to