Elena Stepanova added a comment - 2016-11-18 12:55 InnoDB does not have ROW_FORMAT FIXED . It didn't have it in MySQL 5.6 either. However, in InnoDB 5.6 (both in MySQL 5.6 and MariaDB 10.0/10.1) innodb_strict_mode=0 by default, that's why the CREATE statement simply issued a warning and assumed COMPACT format instead. MySQL [test]> select @@version; + --------------+ | @@version | + --------------+ | 5.6.33-debug | + --------------+ 1 row in set (0.00 sec)   MySQL [test]> CREATE TABLE `__syssequences` ( -> `systemName` CHAR (50) NOT NULL , -> `sequenceName` CHAR (50) NOT NULL , -> `sequenceStep` INT (11) DEFAULT '1' , -> `lastValue` INT (11) DEFAULT '1' , -> `lastInsert` DATETIME DEFAULT '0000-00-00 00:00:00' , -> `lastUpdate` DATETIME DEFAULT '0000-00-00 00:00:00' , -> PRIMARY KEY (`systemName`,`sequenceName`) -> ) ENGINE=INNODB DEFAULT CHARSET=utf8 ROW_FORMAT=FIXED; Query OK, 0 rows affected, 1 warning (0.35 sec)   MySQL [test]> show warnings; + ---------+------+--------------------------------------+ | Level | Code | Message | + ---------+------+--------------------------------------+ | Warning | 1478 | InnoDB: assuming ROW_FORMAT=COMPACT. | + ---------+------+--------------------------------------+ 1 row in set (0.00 sec) In InnoDB 5.7 (both in MySQL 5.7 and MariaDB 10.2) innodb_strict_mode=1 by default, so both of them fail to create a table with the wrong ROW_FORMAT. If you want the old behavior, set innodb_strict_mode=0 . greenman , Maybe it should be explained somehow in the documentation if it hasn't been yet (I couldn't find). While technically every part of it is documented separately – supported row formats as well as the new default for the strict mode, – it might be not easy for users to put it all together. MySQL, for example, describes it directly at CREATE TABLE page: https://dev.mysql.com/doc/refman/5.7/en/create-table.html

Ian Gilfillan added a comment - 2016-11-19 16:21 I have added some detail to the CREATE TABLE page, and, probably more importantly, highlighted the strict mode change on the 10.2.2 release notes page. Also discovered the rather obscure bug I reported at https://jira.mariadb.org/browse/MDEV-11316 as a result. Closed as not a bug, as the behaviour described in the original report is expected with the change to the default strict mode.