database/mariadb
1
0
Fork 0
mirror of https://github.com/MariaDB/server.git synced 2025-08-08 11:22:35 +03:00
Code Activity

MDEV-7118: Anemometer stop working after upgrade to from...

Browse Source 
When the optimizer considers an option to use Loose Scan, it should 
still consider UNIQUE keys (Previously, MDEV-4120 disabled loose scan
for all kinds of unique indexes. That was wrong)

However, we should not use Loose Scan when trying to satisfy 
 "SELECT DISTINCT col1, col2, .. colN"
when using an index defined as UNIQU(col1, col2, ... colN).
This commit is contained in:
Sergey Petrunya
2014-11-19 17:14:49 +03:00
parent 416f267a7a
commit 00475d40d1
4 changed files with 19 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
sql/opt_range.cc
View File

@@ -12801,11 +12801,11 @@ get_best_group_min_max(PARAM *param, SEL_TREE *tree, double read_time)
uint cur_used_key_parts;
/*
Check (B1) - if current index is covering. Exclude UNIQUE indexes, because
loose scan may still be chosen for them due to imperfect cost calculations.
Check (B1) - if current index is covering.
(was also: "Exclude UNIQUE indexes ..." but this was removed because
there are cases Loose Scan over a multi-part index is useful).
*/
if (!table->covering_keys.is_set(cur_index) ||
cur_index_info->flags & HA_NOSAME)
if (!table->covering_keys.is_set(cur_index))
goto next_index;
/*
@@ -12943,6 +12943,16 @@ get_best_group_min_max(PARAM *param, SEL_TREE *tree, double read_time)
min_max_arg_part= cur_index_info->key_part + key_part_nr - 1;
}
/*
Aplly a heuristic: there is no point to use loose index scan when we're
using the whole unique index.
*/
if (cur_index_info->flags & HA_NOSAME &&
cur_group_key_parts == cur_index_info->user_defined_key_parts)
{
goto next_index;
}
/*
Check (NGA1, NGA2) and extract a sequence of constants to be used as part
of all search keys.