leaves12138 opened a new pull request, #8882:
URL: https://github.com/apache/paimon/pull/8882
### Purpose
Reduce peak heap usage during row ID reassignment planning by avoiding full
`ManifestEntry` materialization.
### Changes
- Read only the manifest fields required by assignment planning.
- Intern partition values and retain each active ADD entry in three primitive
words: partition reference, first row ID, and row count.
- Store logical ranges and row ID mappings in primitive arrays.
- Preserve legacy partition ordering, DELETE matching, and generated
assignment
plans.
- Keep the previous planner as a test-only differential oracle.
### Tests
Run on Linux x86_64 against the latest `master`:
- `DataEvolutionRowIdReassignerTest`: 37 passed
- `DataEvolutionRowIdReassignerCompactPlanTest`: 8 passed
- `RowRangeMappingIndexTest`: 10 passed
Total: 55 passed, 0 failures, 0 errors, 0 skipped.
--
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]
