Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Fixed
-
10.0.9
-
None
Description
After fix for MDEV-5926, we see a regression for the filtered% column for testcase from MDEV-4350:
create table t1 (a int);
|
insert into t1 values (1), (1);
|
insert into t1 select * from t1;
|
insert into t1 select * from t1;
|
insert into t1 select * from t1;
|
insert into t1 select * from t1;
|
insert into t1 select * from t1;
|
insert into t1 select * from t1;
|
insert into t1 select * from t1;
|
insert into t1 select * from t1;
|
insert into t1 select * from t1;
|
insert into t1 values (0);
|
set use_stat_tables='preferably';
|
set histogram_size=127;
|
set histogram_type='SINGLE_PREC_HB';
|
analyze table t1;
|
flush table t1;
|
set optimizer_use_condition_selectivity=4;
|
mysql> explain extended select * from t1 where a=0;
|
+------+-------------+-------+------+---------------+------+---------+------+------+----------+-------------+
|
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | filtered | Extra |
|
+------+-------------+-------+------+---------------+------+---------+------+------+----------+-------------+
|
| 1 | SIMPLE | t1 | ALL | NULL | NULL | NULL | NULL | 1025 | 99.22 | Using where |
|
+------+-------------+-------+------+---------------+------+---------+------+------+----------+-------------+
|
filtered% used to be 50%, now it's 99.22%. Both estimates are very wrong:
mysql> select a,count(*) from t1 group by a;
|
+------+----------+
|
| a | count(*) |
|
+------+----------+
|
| 0 | 1 |
|
| 1 | 1024 |
|
+------+----------+
|
But the new one is even worse than before.
Attachments
Issue Links
- relates to
-
-
- Closed
-