hudi-agent commented on code in PR #18611: URL: https://github.com/apache/hudi/pull/18611#discussion_r3532980816
########## hudi-client/hudi-client-common/src/main/java/org/apache/hudi/metadata/HoodieBackedTableMetadataWriter.java: ########## Review Comment: 🤖 **Line 1079:** This is still shared with the restore-sync path: `listAllPartitionsFromFilesystem` is also called from `update(HoodieRestoreMetadata, ...)` at line 1744, so both the filtering and the new `skipped_zero_size_files_on_initialize` metric fire during restore, which isn't an initialize. In that path a zero-size file that's present on disk and tracked in MDT gets excluded from the FS listing and lands in `filesDeleted` (lines 2320-2321), producing a spurious MDT delete while the file still exists. Would passing `skipZeroSizeFiles` as an explicit parameter — `true` from the initialize caller at line 431, `false` from restore at line 1744 — keep this scoped as the PR describes? <sub><i>⚠️ AI-generated; verify before applying. React 👍/👎 to flag quality.</i></sub> -- 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]
