tzphh opened a new pull request, #10459: URL: https://github.com/apache/paimon/pull/10459
Fix #10458. Data Evolution compaction drains a bin when its weight exceeds the target file size, but then discards it if the file count is below `compaction.min.file-num`. For example, 12 contiguous 200 MiB files with a 512 MiB target and a minimum file count of 5 produce no compaction tasks. Allow normal-file bins with multiple files to compact when either the file-count threshold is met or their weight exceeds the target size. Apply the same condition in candidate range collection and task planning. Keep the existing partition, row-ID continuity, single-file splitting, and completed-file checks. ### Tests Add regression coverage for the reported case, exact and exceeded size thresholds, small tails, open-file cost, overlapping column versions, and row-ID gaps. A Parquet test performs compaction and commit, then verifies file count, row IDs, and row values. -- 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]
