You are viewing a plain text version of this content. The canonical link for it is here.

Posted to commits@hudi.apache.org by "nsivabalan (via GitHub)" <gi...@apache.org> on 2026/06/25 03:45:07 UTC

Re: [I] No precombine key support for Spark write/read [hudi]

nsivabalan commented on issue #16658:
URL: https://github.com/apache/hudi/issues/16658#issuecomment-4795628503

   Triage: this looks resolved on current `master`.
   
   - **HUDI-8413** (#12600, commit `4bea55a64415`) changed the default merge mode to `COMMIT_TIME_ORDERING`, which orders by commit/processing time and requires no precombine field — exactly the proc_time-sequence behavior described here.
   - **HUDI-8551** (#12292, commit `8f6b1aed3ef7`) added explicit no-precombine support for `EVENT_TIME_ORDERING`: `HoodieAvroIndexedRecord.doGetOrderingValue` and `HoodieSparkRecord.doGetOrderingValue` now return `OrderingValues.getDefault()` when `orderingFields` is null/empty.
   - `HoodiePayloadConfig.ORDERING_FIELDS` now uses `noDefaultValue()` — the old silent `"ts"` default is gone, so users no longer have to opt out of it.
   - `TestMORDataSource` has parameterized cases for `EVENT_TIME_ORDERING without precombine field` covering both AVRO and SPARK record types.
   
   If there's still a specific Spark write/read scenario where no-precombine doesn't work, could you share a minimal repro? Otherwise this can likely be closed.


-- 
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: commits-unsubscribe@hudi.apache.org

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