MySQL中InnoDB引擎对索引的扩展

原文链接:https://www.modb.pro/db/22927 (复制至浏览器,即可查看)

摘要:InnoDB引擎对索引的扩展,自动追加主键值及其对执行计划的影响。


MySQL中,使用InnoDB引擎的每个表,创建的普通索引(即非主键索引),都会同时保存主键的值。

比如语句

<code>CREATE TABLE t1 (
i1 INT NOT NULL DEFAULT 0,
i2 INT NOT NULL DEFAULT 0,
d DATE DEFAULT NULL,
PRIMARY KEY (i1, i2),
INDEX k_d (d)
) ENGINE = InnoDB;/<code>

创建了t1表,其主键为(i1, i2),同时创建了基于d列的索引k_d,但其实在底层,InnoDB引擎将索引k_d扩展成(d,i1,i2)。

InnoDB引擎这么做,是用空间换性能,优化器在判断是否使用索引及使用哪个索引时会有更多列参考,这样可能生成更高效的执行计划,获得更好的性能。

优化器在ref、range和index_merge类型的访问,Loose Index Scan访问,连接和排序优化, MIN()/MAX()优化时使都会使用扩展列。

我们来看个例子:

<code>root@database-one 15:15:  [gftest]> CREATE TABLE t1 (
-> i1 INT NOT NULL DEFAULT 0,

-> i2 INT NOT NULL DEFAULT 0,
-> d DATE DEFAULT NULL,
-> PRIMARY KEY (i1, i2),
-> INDEX k_d (d)
-> ) ENGINE = InnoDB;
Query OK, 0 rows affected (0.06 sec)

root@database-one 15:15: [gftest]> INSERT INTO t1 VALUES
-> (1, 1, '1998-01-01'), (1, 2, '1999-01-01'),
-> (1, 3, '2000-01-01'), (1, 4, '2001-01-01'),
-> (1, 5, '2002-01-01'), (2, 1, '1998-01-01'),
-> (2, 2, '1999-01-01'), (2, 3, '2000-01-01'),
-> (2, 4, '2001-01-01'), (2, 5, '2002-01-01'),
-> (3, 1, '1998-01-01'), (3, 2, '1999-01-01'),
-> (3, 3, '2000-01-01'), (3, 4, '2001-01-01'),
-> (3, 5, '2002-01-01'), (4, 1, '1998-01-01'),
-> (4, 2, '1999-01-01'), (4, 3, '2000-01-01'),
-> (4, 4, '2001-01-01'), (4, 5, '2002-01-01'),
-> (5, 1, '1998-01-01'), (5, 2, '1999-01-01'),
-> (5, 3, '2000-01-01'), (5, 4, '2001-01-01'),
-> (5, 5, '2002-01-01');
Query OK, 25 rows affected (0.01 sec)
Records: 25 Duplicates: 0 Warnings: 0

root@database-one 15:21: [gftest]> show index from t1;
+-------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| Table | Non_unique | Key_name | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | Index_comment |
+-------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| t1 | 0 | PRIMARY | 1 | i1 | A | 5 | NULL | NULL | | BTREE | | |
| t1 | 0 | PRIMARY | 2 | i2 | A | 25 | NULL | NULL | | BTREE | | |
| t1 | 1 | k_d | 1 | d | A | 5 | NULL | NULL | YES | BTREE | | |
+-------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
3 rows in set (0.01 sec)/<code>

在普通索引中追加扩展主键是InnoDB在底层做的,show index等语句不显示追加列,但我们可以通过其它方式来验证。看这个SQL

SELECT COUNT(*) FROM t1 WHERE i1 = 3 AND d = ‘2000-01-01’

如果InnoDB没有扩展索引,索引k_d为(d),生成的执行计划应该类似这样,使用k_d索引找到d为’2000-01-01’的5行数据,再回表过滤出i1为3的,最后计算count。或者使用主键索引找到i1为3的5行数据,再回表过滤出d为’2000-01-01’的,最后计算count。下面仅示意走k_d索引的情况:

<code>mysql> EXPLAIN SELECT COUNT(*) FROM t1 WHERE i1 = 3 AND d = '2000-01-01'\\G
*************************** 1. row ***************************
id: 1
select_type: SIMPLE
table: t1
type: ref
possible_keys: PRIMARY,k_d
key: k_d
key_len: 4
ref: const
rows: 5
Extra: Using where; Using index/<code>

如果InnoDB扩展了索引,索引k_d为(d,i1,i2),这时,优化器可以使用最左边的索引前缀(d,i1),生成的执行计划应该类似这样,使用k_d索引找到d为’2000-01-01’及i1为3的1行数据,然后计算count

<code>mysql> EXPLAIN SELECT COUNT(*) FROM t1 WHERE i1 = 3 AND d = '2000-01-01'\\G
*************************** 1. row ***************************
id: 1
select_type: SIMPLE
table: t1
type: ref
possible_keys: PRIMARY,k_d
key: k_d
key_len: 8
ref: const,const
rows: 1
Extra: Using index/<code>

并且d列是DATE类型占4个字节,i1是INT类型占4个字节,所以查询中使用的键值长度就是8个字节(key_len: 8)。

我们看看实际生成的执行计划

<code>root@database-one 15:35:  [gftest]> EXPLAIN SELECT COUNT(*) FROM t1 WHERE i1 = 3 AND d = '2000-01-01'\\G
*************************** 1. row ***************************
id: 1
select_type: SIMPLE
table: t1

partitions: NULL
type: ref
possible_keys: PRIMARY,k_d
key: k_d
key_len: 8
ref: const,const
rows: 1
filtered: 100.00
Extra: Using index
1 row in set, 1 warning (0.01 sec)/<code>

果然跟我们的判断一致,注意执行计划中的细节:

  • key_len从4字节变为8字节,表明键查找使用列d和i1,而不仅仅是d。
  • ref从const更改为const,const,表明查找使用两个键值,而不是一个。
  • rows从5减少到1,表明检索更少的行。
  • Extra从Using where; Using index改为Using index,表示只用索引读取,不必回表。


InnoDB引擎底层扩展普通索引的情况,也可以通过跟MyISAM引擎对比来进行旁证:

<code>root@database-one 16:07:  [gftest]> CREATE TABLE t1MyISAM (
-> i1 INT NOT NULL DEFAULT 0,
-> i2 INT NOT NULL DEFAULT 0,
-> d DATE DEFAULT NULL,
-> PRIMARY KEY (i1, i2),
-> INDEX k_d (d)
-> ) ENGINE = MyISAM;
Query OK, 0 rows affected (0.01 sec)

root@database-one 16:07: [gftest]> INSERT INTO t1myisam VALUES
-> (1, 1, '1998-01-01'), (1, 2, '1999-01-01'),
-> (1, 3, '2000-01-01'), (1, 4, '2001-01-01'),
-> (1, 5, '2002-01-01'), (2, 1, '1998-01-01'),
-> (2, 2, '1999-01-01'), (2, 3, '2000-01-01'),
-> (2, 4, '2001-01-01'), (2, 5, '2002-01-01'),
-> (3, 1, '1998-01-01'), (3, 2, '1999-01-01'),
-> (3, 3, '2000-01-01'), (3, 4, '2001-01-01'),
-> (3, 5, '2002-01-01'), (4, 1, '1998-01-01'),
-> (4, 2, '1999-01-01'), (4, 3, '2000-01-01'),
-> (4, 4, '2001-01-01'), (4, 5, '2002-01-01'),
-> (5, 1, '1998-01-01'), (5, 2, '1999-01-01'),
-> (5, 3, '2000-01-01'), (5, 4, '2001-01-01'),
-> (5, 5, '2002-01-01');
Query OK, 25 rows affected (0.02 sec)
Records: 25 Duplicates: 0 Warnings: 0

root@database-one 16:07: [gftest]> EXPLAIN SELECT COUNT(*) FROM t1myisam WHERE i1 = 3 AND d = '2000-01-01'\\G
*************************** 1. row ***************************
id: 1
select_type: SIMPLE
table: t1myisam
partitions: NULL
type: ref
possible_keys: PRIMARY,k_d
key: PRIMARY
key_len: 4
ref: const
rows: 4
filtered: 16.00
Extra: Using where
1 row in set, 1 warning (0.01 sec)/<code>

可以看到,同样的结构同样的数据,因为MyISAM引擎不会在底层自动扩展普通索引,所以执行计划还是通过主键索引进行处理。

按照官方手册的说明,也可以用SHOW STATUS命令来验证

<code>root@database-one 16:12:  [gftest]> FLUSH TABLE t1;
Query OK, 0 rows affected (0.00 sec)

root@database-one 16:12: [gftest]> FLUSH STATUS;
Query OK, 0 rows affected (0.14 sec)

root@database-one 16:12: [gftest]> SELECT COUNT(*) FROM t1 WHERE i1 = 3 AND d = '2000-01-01';
+----------+
| COUNT(*) |
+----------+
| 1 |
+----------+
1 row in set (0.03 sec)

root@database-one 16:12: [gftest]> SHOW STATUS LIKE 'handler_read%';
+-----------------------+-------+
| Variable_name | Value |
+-----------------------+-------+
| Handler_read_first | 0 |
| Handler_read_key | 1 |
| Handler_read_last | 0 |
| Handler_read_next | 1 |
| Handler_read_prev | 0 |
| Handler_read_rnd | 0 |
| Handler_read_rnd_next | 0 |
+-----------------------+-------+
7 rows in set (0.01 sec)

root@database-one 16:13: [gftest]> FLUSH TABLE t1myisam;
Query OK, 0 rows affected (0.01 sec)

root@database-one 16:13: [gftest]> FLUSH STATUS;
Query OK, 0 rows affected (0.00 sec)

root@database-one 16:13: [gftest]> SELECT COUNT(*) FROM t1myisam WHERE i1 = 3 AND d = '2000-01-01';
+----------+
| COUNT(*) |
+----------+
| 1 |
+----------+
1 row in set (0.01 sec)

root@database-one 16:13: [gftest]> SHOW STATUS LIKE 'handler_read%';
+-----------------------+-------+
| Variable_name | Value |
+-----------------------+-------+
| Handler_read_first | 0 |
| Handler_read_key | 1 |
| Handler_read_last | 0 |
| Handler_read_next | 5 |
| Handler_read_prev | 0 |
| Handler_read_rnd | 0 |
| Handler_read_rnd_next | 0 |
+-----------------------+-------+
7 rows in set (0.00 sec)/<code>

Handler_read_next表示在进行索引扫描时,按照索引从数据文件里取数据的次数。使用MyISAM引擎的t1myisam表,Handler_read_next值为5,使用InnoDB引擎的t1表,Handler_read_next值减小到1,就是因为InnoDB引擎对索引进行了主键扩展,读取的次数少,效率更好。

默认情况下,优化器分析InnoDB表的索引时会考虑扩展列,但如果因为特殊原因让优化器不考虑扩展列,可以使用SET optimizer_switch = 'use_index_extensions=off’设置。

<code>root@database-one 16:26:  [gftest]> SET optimizer_switch = 'use_index_extensions=off';
Query OK, 0 rows affected (0.01 sec)

root@database-one 16:26: [gftest]> EXPLAIN SELECT COUNT(*) FROM t1 WHERE i1 = 3 AND d = '2000-01-01'\\G
*************************** 1. row ***************************
id: 1
select_type: SIMPLE
table: t1
partitions: NULL
type: ref
possible_keys: PRIMARY,k_d
key: PRIMARY
key_len: 4
ref: const
rows: 5
filtered: 20.00
Extra: Using where
1 row in set, 1 warning (0.02 sec)/<code>


分享到:


相關文章: