mirror of
https://github.com/MariaDB/server.git
synced 2025-02-12 00:15:35 +01:00
727491b72a
This includes all test changes from
"Changing all cost calculation to be given in milliseconds"
and forwards.
Some of the things that caused changes in the result files:
- As part of fixing tests, I added 'echo' to some comments to be able to
easier find out where things where wrong.
- MATERIALIZED has now a higher cost compared to X than before. Because
of this some MATERIALIZED types have changed to DEPENDEND SUBQUERY.
- Some test cases that required MATERIALIZED to repeat a bug was
changed by adding more rows to force MATERIALIZED to happen.
- 'Filtered' in SHOW EXPLAIN has in many case changed from 100.00 to
something smaller. This is because now filtered also takes into
account the smallest possible ref access and filters, even if they
where not used. Another reason for 'Filtered' being smaller is that
we now also take into account implicit filtering done for subqueries
using FIRSTMATCH.
(main.subselect_no_exists_to_in)
This is caluculated in best_access_path() and stored in records_out.
- Table orders has changed because more accurate costs.
- 'index' and 'ALL' for small tables has changed to use 'range' or
'ref' because of optimizer_scan_setup_cost.
- index can be changed to 'range' as 'range' optimizer assumes we don't
have to read the blocks from disk that range optimizer has already read.
This can be confusing in the case where there is no obvious where clause
but instead there is a hidden 'key_column > NULL' added by the optimizer.
(main.subselect_no_exists_to_in)
- Scan on primary clustered key does not report 'Using Index' anymore
(It's a table scan, not an index scan).
- For derived tables, the number of rows is now 100 instead of 2,
which can be seen in EXPLAIN.
- More tests have "Using index for group by" as the cost of this
optimization is now more correct (lower).
- A primary key could be preferred for a normal key, even if it would
access more rows, as it's faster to do 1 lokoup and 3 'index_next' on a
clustered primary key than one lookup trough a secondary.
(main.stat_tables_innodb)
Notes:
- There was a 4.7% more calls to best_extension_by_limited_search() in
the main.greedy_optimizer test. However examining the test results
it looked that the plans where slightly better (eq_ref where more
chained together) so I assume this is ok.
- I have verified a few test cases where there was notable/unexpected
changes in the plan and in all cases the new optimizer plans where
faster. (main.greedy_optimizer and some others)
196 lines
6.6 KiB
Text
196 lines
6.6 KiB
Text
UPDATE performance_schema.setup_instruments SET enabled = 'NO', timed = 'YES';
|
|
UPDATE performance_schema.setup_instruments SET enabled = 'YES'
|
|
WHERE name in ('wait/io/table/sql/handler',
|
|
'wait/lock/table/sql/handler');
|
|
drop procedure if exists before_payload;
|
|
drop procedure if exists after_payload;
|
|
create procedure before_payload()
|
|
begin
|
|
TRUNCATE TABLE performance_schema.table_io_waits_summary_by_index_usage;
|
|
TRUNCATE TABLE performance_schema.table_io_waits_summary_by_table;
|
|
TRUNCATE TABLE performance_schema.events_waits_history_long;
|
|
TRUNCATE TABLE performance_schema.events_waits_history;
|
|
TRUNCATE TABLE performance_schema.events_waits_current;
|
|
end
|
|
$$
|
|
create procedure after_payload()
|
|
begin
|
|
select count(1) as number_seen,
|
|
OBJECT_TYPE, OBJECT_SCHEMA, OBJECT_NAME, INDEX_NAME,
|
|
OPERATION, NUMBER_OF_BYTES
|
|
from performance_schema.events_waits_history_long
|
|
where OBJECT_SCHEMA = "test"
|
|
group by OBJECT_TYPE, OBJECT_SCHEMA, OBJECT_NAME, INDEX_NAME, OPERATION, NUMBER_OF_BYTES;
|
|
select OBJECT_TYPE, OBJECT_SCHEMA, OBJECT_NAME, INDEX_NAME,
|
|
COUNT_STAR, COUNT_READ, COUNT_WRITE
|
|
from performance_schema.table_io_waits_summary_by_index_usage
|
|
where OBJECT_SCHEMA = "test"
|
|
order by OBJECT_TYPE, OBJECT_SCHEMA, OBJECT_NAME, INDEX_NAME;
|
|
select OBJECT_TYPE, OBJECT_SCHEMA, OBJECT_NAME,
|
|
COUNT_STAR, COUNT_READ, COUNT_WRITE
|
|
from performance_schema.table_io_waits_summary_by_table
|
|
where OBJECT_SCHEMA = "test"
|
|
order by OBJECT_TYPE, OBJECT_SCHEMA, OBJECT_NAME;
|
|
end
|
|
$$
|
|
drop table if exists t1;
|
|
drop table if exists t2;
|
|
drop table if exists t3;
|
|
create table t0(v int);
|
|
create table t1(id1 int, a int);
|
|
create table t2(id1 int, id2 int, b int);
|
|
create table t3(id2 int, id3 int, c int);
|
|
insert into t0 values
|
|
(0), (1), (2), (3), (4),
|
|
(5), (6), (7), (8), (9);
|
|
insert into t1(id1, a)
|
|
select v, 100*v from t0;
|
|
insert into t2(id1, id2, b)
|
|
select X.v, 10*X.v + Y.v, 100*X.v + 10*Y.v
|
|
from t0 X, t0 Y;
|
|
insert into t3(id2, id3, c)
|
|
select 10*X.v + Y.v, 100*X.v + 10*Y.v + Z.v, 100*X.v + 10*Y.v + Z.v
|
|
from t0 X, t0 Y, t0 Z;
|
|
analyze table t1;
|
|
Table Op Msg_type Msg_text
|
|
test.t1 analyze status Engine-independent statistics collected
|
|
test.t1 analyze status OK
|
|
analyze table t2;
|
|
Table Op Msg_type Msg_text
|
|
test.t2 analyze status Engine-independent statistics collected
|
|
test.t2 analyze status OK
|
|
analyze table t3;
|
|
Table Op Msg_type Msg_text
|
|
test.t3 analyze status Engine-independent statistics collected
|
|
test.t3 analyze status OK
|
|
select * from t1 order by a;
|
|
id1 a
|
|
0 0
|
|
1 100
|
|
2 200
|
|
3 300
|
|
4 400
|
|
5 500
|
|
6 600
|
|
7 700
|
|
8 800
|
|
9 900
|
|
select * from t2
|
|
where (b >= 180) and (b <= 220)
|
|
order by b;
|
|
id1 id2 b
|
|
1 18 180
|
|
1 19 190
|
|
2 20 200
|
|
2 21 210
|
|
2 22 220
|
|
select * from t3
|
|
where (c >= 587) and (c <= 612)
|
|
order by c;
|
|
id2 id3 c
|
|
58 587 587
|
|
58 588 588
|
|
58 589 589
|
|
59 590 590
|
|
59 591 591
|
|
59 592 592
|
|
59 593 593
|
|
59 594 594
|
|
59 595 595
|
|
59 596 596
|
|
59 597 597
|
|
59 598 598
|
|
59 599 599
|
|
60 600 600
|
|
60 601 601
|
|
60 602 602
|
|
60 603 603
|
|
60 604 604
|
|
60 605 605
|
|
60 606 606
|
|
60 607 607
|
|
60 608 608
|
|
60 609 609
|
|
61 610 610
|
|
61 611 611
|
|
61 612 612
|
|
explain extended select t1.*, t2.*, t3.*
|
|
from t1 join t2 using (id1) join t3 using (id2);
|
|
id select_type table type possible_keys key key_len ref rows filtered Extra
|
|
1 SIMPLE t1 ALL NULL NULL NULL NULL 10 100.00
|
|
1 SIMPLE t2 ALL NULL NULL NULL NULL 100 100.00 Using where; Using join buffer (flat, BNL join)
|
|
1 SIMPLE t3 ALL NULL NULL NULL NULL 1000 100.00 Using where; Using join buffer (incremental, BNL join)
|
|
Warnings:
|
|
Note 1003 select `test`.`t1`.`id1` AS `id1`,`test`.`t1`.`a` AS `a`,`test`.`t2`.`id1` AS `id1`,`test`.`t2`.`id2` AS `id2`,`test`.`t2`.`b` AS `b`,`test`.`t3`.`id2` AS `id2`,`test`.`t3`.`id3` AS `id3`,`test`.`t3`.`c` AS `c` from `test`.`t1` join `test`.`t2` join `test`.`t3` where `test`.`t3`.`id2` = `test`.`t2`.`id2` and `test`.`t2`.`id1` = `test`.`t1`.`id1`
|
|
call before_payload();
|
|
select t1.*, t2.*, t3.*
|
|
from t1 join t2 using (id1) join t3 using (id2);
|
|
call after_payload();
|
|
number_seen OBJECT_TYPE OBJECT_SCHEMA OBJECT_NAME INDEX_NAME OPERATION NUMBER_OF_BYTES
|
|
11 TABLE test t1 NULL fetch 1
|
|
1 TABLE test t1 NULL read external NULL
|
|
1 TABLE test t1 NULL read normal NULL
|
|
101 TABLE test t2 NULL fetch 1
|
|
1 TABLE test t2 NULL read external NULL
|
|
1 TABLE test t2 NULL read normal NULL
|
|
1 TABLE test t3 NULL fetch 1000
|
|
1 TABLE test t3 NULL read external NULL
|
|
1 TABLE test t3 NULL read normal NULL
|
|
OBJECT_TYPE OBJECT_SCHEMA OBJECT_NAME INDEX_NAME COUNT_STAR COUNT_READ COUNT_WRITE
|
|
TABLE test t0 NULL 0 0 0
|
|
TABLE test t1 NULL 11 11 0
|
|
TABLE test t2 NULL 101 101 0
|
|
TABLE test t3 NULL 1000 1000 0
|
|
OBJECT_TYPE OBJECT_SCHEMA OBJECT_NAME COUNT_STAR COUNT_READ COUNT_WRITE
|
|
TABLE test t0 0 0 0
|
|
TABLE test t1 11 11 0
|
|
TABLE test t2 101 101 0
|
|
TABLE test t3 1000 1000 0
|
|
alter table t1 add unique index(id1);
|
|
alter table t2 add unique index(id2);
|
|
alter table t2 add index(id1);
|
|
alter table t3 add unique index(id3);
|
|
alter table t3 add index(id2);
|
|
explain extended select t1.*, t2.*, t3.*
|
|
from t1 join t2 using (id1) join t3 using (id2);
|
|
id select_type table type possible_keys key key_len ref rows filtered Extra
|
|
1 SIMPLE t1 ALL id1 NULL NULL NULL 10 100.00
|
|
1 SIMPLE t2 ALL id2,id1 NULL NULL NULL 100 10.00 Using where; Using join buffer (flat, BNL join)
|
|
1 SIMPLE t3 ref id2 id2 5 test.t2.id2 10 100.00
|
|
Warnings:
|
|
Note 1003 select `test`.`t1`.`id1` AS `id1`,`test`.`t1`.`a` AS `a`,`test`.`t2`.`id1` AS `id1`,`test`.`t2`.`id2` AS `id2`,`test`.`t2`.`b` AS `b`,`test`.`t3`.`id2` AS `id2`,`test`.`t3`.`id3` AS `id3`,`test`.`t3`.`c` AS `c` from `test`.`t1` join `test`.`t2` join `test`.`t3` where `test`.`t3`.`id2` = `test`.`t2`.`id2` and `test`.`t2`.`id1` = `test`.`t1`.`id1`
|
|
call before_payload();
|
|
select t1.*, t2.*, t3.*
|
|
from t1 join t2 using (id1) join t3 using (id2);
|
|
call after_payload();
|
|
number_seen OBJECT_TYPE OBJECT_SCHEMA OBJECT_NAME INDEX_NAME OPERATION NUMBER_OF_BYTES
|
|
11 TABLE test t1 NULL fetch 1
|
|
1 TABLE test t1 id1 read external NULL
|
|
1 TABLE test t1 id1 read normal NULL
|
|
101 TABLE test t2 NULL fetch 1
|
|
1 TABLE test t2 id2 read external NULL
|
|
1 TABLE test t2 id2 read normal NULL
|
|
100 TABLE test t3 id2 fetch 10
|
|
1 TABLE test t3 id3 read external NULL
|
|
1 TABLE test t3 id3 read normal NULL
|
|
OBJECT_TYPE OBJECT_SCHEMA OBJECT_NAME INDEX_NAME COUNT_STAR COUNT_READ COUNT_WRITE
|
|
TABLE test t0 NULL 0 0 0
|
|
TABLE test t1 NULL 11 11 0
|
|
TABLE test t1 id1 0 0 0
|
|
TABLE test t2 NULL 101 101 0
|
|
TABLE test t2 id1 0 0 0
|
|
TABLE test t2 id2 0 0 0
|
|
TABLE test t3 id2 1000 1000 0
|
|
TABLE test t3 id3 0 0 0
|
|
OBJECT_TYPE OBJECT_SCHEMA OBJECT_NAME COUNT_STAR COUNT_READ COUNT_WRITE
|
|
TABLE test t0 0 0 0
|
|
TABLE test t1 11 11 0
|
|
TABLE test t2 101 101 0
|
|
TABLE test t3 1000 1000 0
|
|
drop table t0;
|
|
drop table t1;
|
|
drop table t2;
|
|
drop table t3;
|
|
drop procedure before_payload;
|
|
drop procedure after_payload;
|
|
UPDATE performance_schema.setup_instruments SET enabled = 'YES', timed = 'YES';
|