cshuo opened a new pull request, #19630:
URL: https://github.com/apache/hudi/pull/19630

   ### Describe the issue this Pull Request addresses
   
   Closes #19629.
   
   Flink native Parquet log writers currently use 
`hoodie.parquet.max.file.size` as their rollover threshold. As a result, 
`write.log.max.size` does not control native log files and changing the 
native-log size also requires changing the target size of base Parquet files.
   
   ### Summary and Changelog
   
   - Detect native Parquet log paths when building the Flink RowData Parquet 
writer configuration.
   - Use `hoodie.logfile.max.size` for native log files while retaining 
`hoodie.parquet.max.file.size` for base Parquet files and pathless 
output-stream writers.
   - Add coverage that verifies base and native log writers receive their 
respective maximum file sizes.
   - Validate with `mvn -pl hudi-client/hudi-flink-client -am 
-Dtest=TestHoodieRowDataParquetConfigInjector#testNativeLogUsesLogMaxFileSize 
-Dsurefire.failIfNoSpecifiedTests=false -DskipITs -DskipSparkTests 
-DskipScalaTests test` (1 test, 0 failures, 0 errors).
   
   ### Impact
   
   Flink MOR native Parquet logs now honor the existing log-file size 
configuration independently of the base Parquet file-size configuration. This 
avoids unnecessary writer rollovers when the configured log-file limit is 
larger than the base-file limit. No public API or configuration key changes are 
introduced. Flink mini-batch boundaries can still close write-once native log 
files before this maximum is reached.
   
   ### Risk Level
   
   Low. The change is limited to selecting the maximum file size for Flink 
native Parquet log paths, with a focused unit test covering both native log and 
base file behavior.
   
   ### Documentation Update
   
   None.
   
   ### Contributor's checklist
   
   - [ ] Read through [contributor's 
guide](https://hudi.apache.org/contribute/how-to-contribute)
   - [x] Enough context is provided in the sections above
   - [x] Adequate tests were added if applicable
   


-- 
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