Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Fixed
-
10.8.3
Description
From https://github.com/MariaDB/mariadb-docker/issues/438
Phase 4/7: Running 'mysql_fix_privilege_tables'
|
2022-06-11 6:03:58 10 [ERROR] Incorrect definition of table mysql.column_stats: expected column 'hist_type' at position 9 to have type enum('SINGLE_PREC_HB','DOUBLE_PREC_HB','JSON_HB'), found type enum('SINGLE_PREC_HB','DOUBLE_PREC_HB').
|
2022-06-11 6:03:58 10 [ERROR] Incorrect definition of table mysql.column_stats: expected column 'histogram' at position 10 to have type longblob, found type varbinary(255).
|
2022-06-11 6:03:58 10 [ERROR] Incorrect definition of table mysql.column_stats: expected column 'hist_type' at position 9 to have type enum('SINGLE_PREC_HB','DOUBLE_PREC_HB','JSON_HB'), found type enum('SINGLE_PREC_HB','DOUBLE_PREC_HB').
|
2022-06-11 6:03:58 10 [ERROR] Incorrect definition of table mysql.column_stats: expected column 'histogram' at position 10 to have type longblob, found type varbinary(255).
|
2022-06-11 6:04:01 10 [ERROR] Incorrect definition of table mysql.column_stats: expected column 'hist_type' at position 9 to have type enum('SINGLE_PREC_HB','DOUBLE_PREC_HB
|
Confirmed with 10.3.34, 10.5.15 and 10.6.7 as the source version before the upgrade.
Attachments
Issue Links
- relates to
-
-
- Closed
-
Activity
Upon analysis the mariadb-upgrade, despite the very large amount of noise, was actually successful. Because the mysql.column_stats was one of the last table to be corrected, the previous ALTER TABLE sql statements in the mariadb-upgrade all caused errors to be emitted by the server.
Per the output in the mariadb-docker issue "2022-06-11 06:04:11+00:00 [Note] [Entrypoint]: Finished mariadb-upgrade" did occur so it was completed.
A correction is in review https://github.com/MariaDB/server/pull/2166 to make the output less noisy. I'm yet to confirm what makes 10.8 noisy here.
Reviewed upgrade scenarios and commented in https://github.com/MariaDB/server/pull/2166.
Also reviewed GTID number change in test case caused by add_suppression procedure's INSERT statement.
OK to push.
Can confirm this issue for the upgrade from 10.6.5