ujjawal4046 opened a new pull request, #8819: URL: https://github.com/apache/hbase/pull/8819
## Overview When a bulk load HFile spans multiple regions, the tool writes split outputs under the column family's `.tmp` directory and removes the original input. If the load then fails before those outputs are loaded, a later invocation skips `.tmp`, loses the only remaining copies, and cannot recover the load. This change: * discovers split HFiles under each family `.tmp` directory when preparing the load queue; * treats `.top` and `.bottom` files in `.tmp` as complete split HFiles rather than references; * adds a regression test that fails after splitting, retries with a new `BulkLoadHFiles` instance, and verifies the data loads and temporary files are cleaned up. ## Testing * `mvn -pl hbase-server -Dtest=TestBulkLoadHFilesSplitRecovery test` (9 tests passed) * `mvn -pl hbase-server -am -DskipTests install` (build passed) * `mvn -pl hbase-server -DskipTests validate` (passed) -- 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]
