vikramahuja1001 commented on code in PR #5976: URL: https://github.com/apache/hive/pull/5976#discussion_r2232818285
########## ql/src/test/results/clientpositive/llap/msck_repair_9.q.out: ########## @@ -90,7 +90,7 @@ PREHOOK: Output: default@tbl_y POSTHOOK: query: MSCK REPAIR TABLE tbl_y POSTHOOK: type: MSCK POSTHOOK: Output: default@tbl_y -Partitions not in metastore: tbl_y:month=12/day=2 tbl_y:month=12/day=3 tbl_y:month=12/day=__HIVE_DEFAULT_PARTITION__ tbl_y:month=ANOTHER_PARTITION/day=3 +Partitions not in metastore: tbl_y:month=12/day=2 tbl_y:month=12/day=3 tbl_y:month=ANOTHER_PARTITION/day=3 Review Comment: @deniskuzZ, in the above scenario that you described: `INSERT INTO tbl_x (id, name, month) values(3, 'ccc', 12);` -> PARTITION(month=12,day=HIVE_DEFAULT_PARTITION) `SET hive.exec.default.partition.name=ANOTHER_PARTITION;` `INSERT INTO tbl_x (id, name, day) values(4, 'ddd', 3);` -> PARTITION(month=ANOTHER_PARTITION,day=3) `MSCK REPAIR TABLE tbl_x;` hive.exec.default.partition.name is an exactly one value. So at one point of time only one one value is valid, rest all the other values are invalid. So when MSCK repair is ran , `hive.exec.default.partition.name` has been already set to `ANOTHER_PARTITION` which makes the HIVE_DEFAULT_PARTITION not a valid partition anymore. Since by default `hive.msck.path.validation` is set to `throw`, msck repair will throw an exception as it encounters a invalid partition (month=12,day=HIVE_DEFAULT_PARTITION) and will fail the whole operation without adding any new partition to the HMS with the exception: `Invalid partition value: Cannot parse HIVE_DEFAULT_PARTITION as type int. Failing MSCK. Set hive.msck.path.validation=skip to ignore invalid partitions` But when the user sets `hive.msck.path.validation=skip` at the session level, msck repair operation will succeed but it will skip all the invalid partitions by adding a warning log (month=12,day=HIVE_DEFAULT_PARTITION in this case) and will only add the valid partitions in that case((month=ANOTHER_PARTITION,day=3). -- 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: gitbox-unsubscr...@hive.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: gitbox-unsubscr...@hive.apache.org For additional commands, e-mail: gitbox-h...@hive.apache.org
