Hi, Tom! On Sep 07, Tom Ryan wrote:
Sep 6 14:37:59 node-2 mysqld[1846]: /usr/sbin/mysqld(handler_index_cond_check+0x8d)[0x5644f6c7b1fd] Sep 6 14:37:59 node-2 mysqld[1846]: /usr/sbin/mysqld(+0x566623)[0x5644f6973623] Sep 6 14:37:59 node-2 mysqld[1846]: /usr/sbin/mysqld(+0xbb4ba8)[0x5644f6fc1ba8] Sep 6 14:37:59 node-2 mysqld[1846]: /usr/sbin/mysqld(+0xad1cfa)[0x5644f6edecfa] Sep 6 14:37:59 node-2 mysqld[1846]: /usr/sbin/mysqld(+0xad1fa3)[0x5644f6edefa3] Sep 6 14:37:59 node-2 mysqld[1846]: /usr/sbin/mysqld(handler::ha_rnd_next(unsigned char*)+0x12f)[0x5644f6c75acf] Sep 6 14:37:59 node-2 mysqld[1846]: /usr/sbin/mysqld(rr_sequential(READ_RECORD*)+0x22)[0x5644f6d988e2] Sep 6 14:37:59 node-2 mysqld[1846]: /usr/sbin/mysqld(sub_select(JOIN*, st_join_table*, bool)+0x164)[0x5644f6a97034] Sep 6 14:37:59 node-2 mysqld[1846]: /usr/sbin/mysqld(JOIN::exec_inner()+0xc9f)[0x5644f6ab85af] Sep 6 14:37:59 node-2 mysqld[1846]: /usr/sbin/mysqld(JOIN::exec()+0x33)[0x5644f6ab8833] Sep 6 14:37:59 node-2 mysqld[1846]: /usr/sbin/mysqld(mysql_select(THD*, TABLE_LIST*, unsigned int, List<Item>&, Item*, unsigned int, st_order*, st_order*, Item*, st_order*, unsigned long long, select_result*, st_select_lex_unit*, st_select_lex*)+0x186)[0x5644 f6ab6b26] Sep 6 14:37:59 node-2 mysqld[1846]: /usr/sbin/mysqld(handle_select(THD*, LEX*, select_result*, unsigned long)+0xf6)[0x5644f6ab7456] Sep 6 14:37:59 node-2 mysqld[1846]: /usr/sbin/mysqld(+0x55756d)[0x5644f696456d] Sep 6 14:37:59 node-2 mysqld[1846]: /usr/sbin/mysqld(mysql_execute_command(THD*)+0x7371)[0x5644f6a61741]
This looks very similar to https://jira.mariadb.org/browse/MDEV-20252 It's not fixed in 10.4.8, sorry. Comments suggest to use optimizer_switch='rowid_filter=off' as a workaround. Also you can add yourself to the watch list for this bug to get notified when it's fixed. Regards, Sergei VP of MariaDB Server Engineering and security@mariadb.org