MySQL 索引与性能调优

索引用于快速找出在某个列中有一特定值的行,如果不使用索引MySQL必须从第l条记录开始读完整个表,直到找出相关的行.表越大,查询数据所花费的时间越多,如果表中查询的列有一个索引,MySQL能快速到达某个位置去搜寻数据文件,而不必查看所有数据,可加快数据查询的查询速度提高效率,索引可在创建表时增加,也可动态调整已有表.

♥ 文章声明 ♥ 该系列文章部分文字描述,参考于以下文献,化繁为简. 《MySQL5.7从入门到精通》 - 刘增杰

通俗的来说索引是一种数据结构,是帮助MySQL进行高效检索数据的一种机制,你可以简单理解为排好序的快速查找数据结构,

索引都是B+树(多路搜索树)结构组织的索引,包括聚集索引,复合索引,前缀索引,唯一索引,都是b+树索引.

优势：1.提高数据检索效率，降低数据库IO成本，降低CPU消耗。劣势: 2.索引是一张表，索引也占空间，虽然提高了查询速度，但也会降低表的更新速度，如果新加数据，索引也会自动更新。

CPU在饱和的时候一般发生在数据装入内存或从磁盘上读取数据时。 IO磁盘平静，服务器硬件平静。

创建普通索引: 在创建表时指定索引类型,如下在u_id字段添加一个普通索引,该索引作用只是加对快数据的访问速度.

 
MariaDB [lyshark]> create table book
    -> (
    -> u_id int not null,
    -> u_book varchar(20) not null,
    -> index(u_id)
    -> );
 
-- 使用show index语句查看指定表中创建的索引
MariaDB [lyshark]> show index from book;
MariaDB [lyshark]> show create table book \G;
MariaDB [lyshark]> explain select * from book where u_id=1 \G;

创建唯一索引: 创建唯一索引的主要原因是减少查询索引列操作的执行时间,尤其是对比较庞大的数据表.它与前面的普通索引类似,不同的就是:索引列的值必须唯一,但允许有空值.如果是组合索引,则列值的组合必须唯一.

 
MariaDB [lyshark]> create table table_1
    -> (
    -> id int not null,
    -> name char(30) not null,
    -> unique index UniqIdx(id)
    -> );
Query OK, 0 rows affected (0.02 sec)
 
MariaDB [lyshark]> show create table table_1 \G;
*************************** 1. row ***************************
       Table: table_1
Create Table: CREATE TABLE `table_1` (
  `id` int(11) NOT NULL,
  `name` char(30) NOT NULL,
  UNIQUE KEY `UniqIdx` (`id`)             #id字段已经成功建立了一个名为UniqIdx的唯一索引
) ENGINE=InnoDB DEFAULT CHARSET=latin1
1 row in set (0.00 sec)

创建单列索引: 单列索引是在数据表中的某一个字段上创建的索引,一个表中可以创建多个单列索引.前面两个例子中创建的索引都为单列索引.

 
MariaDB [lyshark]> create table table_2
    -> (
    -> id int not null,
    -> name char(50) null,
    -> index SingleIdx(name(20))
    -> );
Query OK, 0 rows affected (0.03 sec)
 
MariaDB [lyshark]> show create table table_2 \G;
*************************** 1. row ***************************
       Table: table_2
Create Table: CREATE TABLE `table_2` (
  `id` int(11) NOT NULL,
  `name` char(50) DEFAULT NULL,
  KEY `SingleIdx` (`name`(20))          #name字段上已经成功建立了一个单列索引,名称为SingleIdx
) ENGINE=InnoDB DEFAULT CHARSET=latin1
1 row in set (0.00 sec)

创建组合索引: 组合索引就是在多个字段上创建多个索引.

 
MariaDB [lyshark]> create table table_3
    -> (
    -> id int not null,
    -> name char(30) not null,
    -> age int not null,
    -> info varchar(255),
    -> index MultiIdx(id,name,age)
    -> );

创建全文索引: 全文索引可以用于全文搜索,全文索引适合用于大型数据集,目前只有MyISAM存储引擎支持FULLTEXT索引,并且只为CHAR、VARCHAR和TEXT列创建索引.索引总是对整个列进行,不支持局部(前缀)索引.

 
-- 注意：MySQL5.7默认存储引擎室是InnoDB,在这里我们要改成MyISAM,不然索引会报错
MariaDB [lyshark]> create table table_4(
    -> id int not null,
    -> name char(40) not null,
    -> age int not null,
    -> info varchar(255),
    -> fulltext index FullTxtIdx(info)
    -> )engine=MyISAM;

创建空间索引: 空间索引必须在MyISAM类型的表中创建,且空间类型的字段必须为空,可以看到,table_5表的g字段上创建了名称为spatIdex的空间索引,注意:创建时间指定空间类型字段值的非空约束,并且表的存储引擎必须为MyISAM.

 
MariaDB [lyshark]> create table table_5
    -> (
    -> g geometry not null,
    -> spatial index spatIdx(g)
    -> )engine=MyISAM;

添加索引: 上面的几种形式都是在新建表中添加索引,如果需要在已存在表中添加则需要使用以下命令了.

 
MariaDB [lyshark]> create table book
    -> (
    -> bookid int not null,
    -> bookname varchar(255) not null,
    -> authors varchar(255) not null,
    -> info varchar(255) null,
    -> comment varchar(255) null,
    -> year_public year not null
    -> );
 
-- 添加普通索引
MariaDB [lyshark]> alter table book add index BKNameIdx(bookname(30));
 
-- 添加唯一索引
MariaDB [lyshark]> alter table book add unique index UniqidIdx(bookId);
 
-- 添加单列索引
MariaDB [lyshark]> alter table book add index BkcmtIdx(comment(50));
 
-- 添加组合索引
MariaDB [lyshark]> alter table book add index BkAuAndInfoIdx(authors(30),info(50));
 
-- 通过索引名字删除索引
MariaDB [lyshark]> alter table book drop index UniqidIdx;
MariaDB [lyshark]> alter table book drop index BKNameIdx;

explain 字段的情况:

 
MariaDB [lyshark]> explain select s_name,s_city from suppliers where s_id IN (select Gid from lyshark where Uid='a1');
+------+-------------+-----------+-------+---------------+---------+---------+-------+------+-------+
| id   | select_type | table     | type  | possible_keys | key     | key_len | ref   | rows | Extra |
+------+-------------+-----------+-------+---------------+---------+---------+-------+------+-------+
|    1 | PRIMARY     | lyshark   | const | PRIMARY       | PRIMARY | 30      | const |    1 |       |
|    1 | PRIMARY     | suppliers | const | PRIMARY       | PRIMARY | 4       | const |    1 |       |
+------+-------------+-----------+-------+---------------+---------+---------+-------+------+-------+
 
MariaDB [lyshark]> explain select Course.CID,Course.Cname from Course join(
select CID from lyshark.StudentScore where SID = (select SID from lyshark.Student where Sname='周梅')
)as StudentScore on Course.CID = StudentScore.CID;
+---------+-------------+--------------+------+---------------+------+---------+------+------+----------+
| id      | select_type | table        | type | possible_keys | key     | key_len | ref  | rows | Extra |
+---------+-------------+--------------+------+---------------+------+---------+------+------+----------+
|    1(1) | SIMPLE      | Course       | ALL  | NULL          | NULL | NULL    | NULL |    3 |          |
|    1(2) | SIMPLE      | StudentScore | ALL  | NULL          | NULL | NULL    | NULL |   18 |          |
|    3    | SUBQUERY    | Student      | ALL  | NULL          | NULL | NULL    | NULL |   10 |          |
+---------+-------------+--------------+------+---------------+------+---------+------+------+----------+
 
ID字段的理解:
1.当ID字段相同的情况下执行数据是从上到下,例如第一张表中会由上至下执行下来.
2.当ID不同的情况,如果是子查询,id越大的将在最前面被执行,例如第二张表执行顺序为3->1(1)->1(2)

select_type

 
MariaDB [lyshark]> explain select * from tbl_emp a left join tbl_dept b on a.deptld=b.id where b.id is null
    -> union
    -> select * from tbl_emp a right join tbl_dept b on a.deptld = b.id where a.deptld is null;
+------+--------------+------------+--------+---------------+---------+---------+------------------+------+---------+
| id   | select_type  | table      | type   | possible_keys | key     | key_len | ref              | rows | Extra   |
+------+--------------+------------+--------+---------------+---------+---------+------------------+------+---------+
|    1 | PRIMARY      | a          | ALL    | NULL          | NULL    | NULL    | NULL             |    8 |         |
|    1 | PRIMARY      | b          | eq_ref | PRIMARY       | PRIMARY | 4       | lyshark.a.deptld |    1 |         |
|    2 | UNION        | b          | ALL    | NULL          | NULL    | NULL    | NULL             |    5 |         |
|    2 | UNION        | a          | ALL    | fk_dept_id    | NULL    | NULL    | NULL             |    8 |         |
| NULL | UNION RESULT | <union1,2> | ALL    | NULL          | NULL    | NULL    | NULL             | NULL |         |
+------+--------------+------------+--------+---------------+---------+---------+------------------+------+---------+
 
SIMPLE       => 标志着简单的select查询请求,查询中不包含子查询或者union查询.
PRIMARY      => 查询中任何复杂的查询中,最外层的查询语句,就是最后加载的语句.
SUBQUERY     => 子查询类型,在select或where列表中包含了子查询.
DERIVED      => 在FROM列表中包含子查询,会被标记为DERIVED(衍生),此时会递归执行子查询,并存储在临时表中.
UNION        => 若第二个SELECT出现在UNION之后,则标记为UNION.
UNION RESULT => 从UNION表中获取结果的SELECT

type 访问类型排列，只要能够达到ref，range级别就已经不错了，性能效率。

 
system -> const -> eq_ref -> ref -> range ->index -> all
 
 
system -> 表中只有一条记录，这是const类型的特里，平时不会出现。
 
const  -> 主键唯一索引：表示通过索引一次就找到数据，例如查询一个常量。
MariaDB [lyshark]> explain select * from lyshark where Uid="a1";
+------+-------------+---------+-------+---------------+---------+---------+-------+------+-------+
| id   | select_type | table   | type  | possible_keys | key     | key_len | ref   | rows | Extra |
+------+-------------+---------+-------+---------------+---------+---------+-------+------+-------+
|    1 | SIMPLE      | lyshark | const | PRIMARY       | PRIMARY | 30      | const |    1 |       |
+------+-------------+---------+-------+---------------+---------+---------+-------+------+-------+
 
 
eq_ref -> 唯一性索引扫描，对于每个索引键，表中只有一条记录与之匹配，常用于主键或唯一索引扫描。
 
ref-> 非唯一性索引扫描，返回匹配某个单独值的所有行，被之上也是一种索引访问。查找扫描混合体
 
MariaDB [lyshark]> create table t1(col1 int,col2 int);
MariaDB [lyshark]> alter table t1 add index idx_col1_col2(col1,col2);
MariaDB [lyshark]> explain select * from t1 where col1=1;
+------+-------------+-------+------+---------------+---------------+---------+-------+------+--------+
| id   | select_type | table | type | possible_keys | key           | key_len | ref   | rows | Extra  |
+------+-------------+-------+------+---------------+---------------+---------+-------+------+--------+
|    1 | SIMPLE      | t1    | ref  | idx_col1_col2 | idx_col1_col2 | 5       | const |    1 |        |
+------+-------------+-------+------+---------------+---------------+---------+-------+------+--------+
 
range -> 范围扫描，只检索给定范围的行，key列显示使用了那个索引。 where，between，<>，in 等查询中使用。
 
explain select * from t1 where col1 between 1 and 2;
explain select * from t1 where col1 in (1,2,3);
 
index -> 全索引扫描，全表索引扫描，比all要好一些。
MariaDB [lyshark]> explain select * from t1;
 
all -> 全表扫描。最差的性能。

possible_keys，key 是否使用到了索引，possible_keys 显示可能

 
possible_keys => 显示可能应用在这张表中的索引，一个或多个，该索引会被列出，但不一定被实际查询使用。
key => 实际应用到本次查询的索引类型。最重要的。如果为NULL，则说明没有使用索引。
 
 
MariaDB [lyshark]> explain select col1,col2 from t1;
+------+-------------+-------+-------+---------------+---------------+---------+------+------+-------+
| id   | select_type | table | type  | possible_keys | key           | key_len | ref  | rows | Extra |
+------+-------------+-------+-------+---------------+---------------+---------+------+------+-------+
|    1 | SIMPLE      | t1    | index | NULL          | idx_col1_col2 | 10      | NULL |    1 |       |
+------+-------------+-------+-------+---------------+---------------+---------+------+------+-------+
 
 
------------------------------------------------------------------------------
查询中若使用了覆盖索引，则该索引仅出现key列表中 覆盖索引以下就是原理
 
MariaDB [lyshark]> create table t1(col1 int,col2 int,col3 int);
MariaDB [lyshark]> alter table t1 add index idx_col1_col2(col1,col2);
 
-- 建立的索引与，查询的行数，必须一致，col1,col2是有索引的。
MariaDB [lyshark]> explain select col1,col2 from t1;
+------+-------------+-------+-------+---------------+---------------+---------+------+------+-------------+
| id   | select_type | table | type  | possible_keys | key           | key_len | ref  | rows | Extra   |
+------+-------------+-------+-------+---------------+---------------+---------+------+------+-------------+
|    1 | SIMPLE      | t1    | index | NULL          | idx_col1_col2 | 10      | NULL |    1 | Using  |
+------+-------------+-------+-------+---------------+---------------+---------+------+------+-------------+
 
-- 扫描三个值，不会出现使用索引的情况。
MariaDB [lyshark]> explain select col1,col2,col3 from t1;
+------+-------------+-------+------+---------------+------+---------+------+------+-------+
| id   | select_type | table | type | possible_keys | key  | key_len | ref  | rows | Extra |
+------+-------------+-------+------+---------------+------+---------+------+------+-------+
|    1 | SIMPLE      | t1    | ALL  | NULL          | NULL | NULL    | NULL |    1 |       |
+------+-------------+-------+------+---------------+------+---------+------+------+-------+

possible_keys，key,ken_len

 
key_len 表示索引中使用的字节数，这个长度用的越少越好，kenLen长度是根据表的定义计算得出，而不是表中数据检索出的。
 
 
 
 
ref 显示索引的那一列被使用了，如果可能的话，是一个常数，那些列或常量被用于查找索引列上的值。定义了引用了那些库。
 
rows 根据表统计信息及索引选用情况，大致估算出找到所需的记录所需要读取的行数，该值当然也是越小越好。 每张表有多少行被优化器查询。
 
MariaDB [lyshark]> explain select * from lyshark;
+------+-------------+---------+------+---------------+------+---------+------+------+-------+
| id   | select_type | table   | type | possible_keys | key  | key_len | ref  | rows | Extra |
+------+-------------+---------+------+---------------+------+---------+------+------+-------+
|    1 | SIMPLE      | lyshark | ALL  | NULL          | NULL | NULL    | NULL |   17 |       |
+------+-------------+---------+------+---------------+------+---------+------+------+-------+
1 row in set (0.00 sec)
 
 
MariaDB [lyshark]> create table tt1(id int primary key,col1 varchar(10),col2 varchar(10));
MariaDB [lyshark]> create table tt2(id int primary key,col1 varchar(10),col2 varchar(10));
MariaDB [lyshark]> create index idt_col1_col2 on tt2(col1,col2);
 
MariaDB [lyshark]> explain select * from tt1,tt2 where tt1.id = tt2.id and tt2.col1 = 'admin';
+------+-------------+-------+--------+-----------------------+---------+---------+----------------+------+-------------+
| id   | select_type | table | type   | possible_keys         | key     | key_len | ref            | rows | Extra       |
+------+-------------+-------+--------+-----------------------+---------+---------+----------------+------+-------------+
|    1 | SIMPLE      | tt1   | ALL    | PRIMARY               | NULL    | NULL    | NULL           |    1 |             |
|    1 | SIMPLE      | tt2   | eq_ref | PRIMARY,idt_col1_col2 | PRIMARY | 4       | lyshark.tt1.id |    1 | Using where |
+------+-------------+-------+--------+-----------------------+---------+---------+----------------+------+-------------+
2 rows in set (0.00 sec)

extra 扩展列

using filesort 产生了文件内排序，完蛋了，mysql无法使用索引进行排序，使用了外部的索引排序，而不是按照表内的索引顺序进行读取。mysql无法利用索引完成排序，操作成为文件排序。

 
MariaDB [lyshark]> create table tab1(id int primary key,col1 int,col2 int,col3 int);
MariaDB [lyshark]> insert into tab1 values(1,1,2,3),(2,4,5,6),(3,7,8,9);
MariaDB [lyshark]> create index tab1_col1_col2_col3 on tab1(col1,col2,col3);
 
MariaDB [lyshark]> explain select col1 from tab1 where col1 order by col3 \G
*************************** 1. row ***************************
           id: 1
  select_type: SIMPLE
        table: tab1
         type: index
possible_keys: tab1_col1_col2_col3
          key: tab1_col1_col2_col3
      key_len: 15
          ref: NULL
         rows: 3
        Extra: Using where; Using index; Using filesort
1 row in set (0.00 sec)
 
MariaDB [lyshark]> explain select col1 from tab1 where col1 order by col2,col3 \G
 
查询使用索引，没问题的，只是说，我们没有针对order by 建立排序索引，或者是建立了索引，你没用上！！
 
 
以下我们加上全部索引字段，从此下面这条sql性能更高了。
MariaDB [lyshark]> explain select col1 from tab1 where col1 order by col1,col2,col3 \G
*************************** 1. row ***************************
           id: 1
  select_type: SIMPLE
        table: tab1
         type: index
possible_keys: tab1_col1_col2_col3
          key: tab1_col1_col2_col3
      key_len: 15
          ref: NULL
         rows: 3
        Extra: Using where; Using index
1 row in set (0.00 sec)
 
 
如果可以，尽快优化。

using temporary 彻底完犊子，这个会新建了一个内部临时表，然后操作完后再把临时表删除，动作更凶险。

使用临时表保存中间结果，mysql在对查询结果排序时使用临时表，常用于排序order by 和分组查询group by .

 
MariaDB [lyshark]> explain select * from tab1 where col1 in(1,2,3) group by col2 \G
*************************** 1. row ***************************
           id: 1
  select_type: SIMPLE
        table: tab1
         type: index
possible_keys: tab1_col1_col2_col3
          key: tab1_col1_col2_col3
      key_len: 15
          ref: NULL
         rows: 3
        Extra: Using where; Using index; Using temporary; Using filesort  彻底完犊子
1 row in set (0.00 sec)
 
解决办法，你给我建立的索引个数和顺序，一定要按顺序来。
MariaDB [lyshark]> explain select * from tab1 where col1 in(1,2,3) group by col1,col2 \G
*************************** 1. row ***************************
           id: 1
  select_type: SIMPLE
        table: tab1
         type: index
possible_keys: tab1_col1_col2_col3
          key: tab1_col1_col2_col3
      key_len: 15
          ref: NULL
         rows: 3
        Extra: Using where; Using index   解决了。
1 row in set (0.00 sec)

using index 这种情况是好事，表示相应的操作使用了 covering index 使用了覆盖索引，效率不错，。

如果同时出现了using where 表示索引被用来执行索引键值的查找。

如果没有同时出现using where 表明索引用来读取数据而非执行查找动作。

 
MariaDB [lyshark]> explain select col1,col2 from tab1;
+------+-------------+-------+-------+---------------+---------------------+---------+------+------+-------------+
| id   | select_type | table | type  | possible_keys | key                 | key_len | ref  | rows | Extra       |
+------+-------------+-------+-------+---------------+---------------------+---------+------+------+-------------+
|    1 | SIMPLE      | tab1  | index | NULL          | tab1_col1_col2_col3 | 15      | NULL |    3 | Using index |  using 表明直接从索引上找到了数据。
+------+-------------+-------+-------+---------------+---------------------+---------+------+------+-------------+
1 row in set (0.00 sec)

覆盖索引，就是说你建立的复合索引是 x,y,z 那么你在查询是应该要覆盖这些索引，这样才能让索引，发挥其最大的性能，否则索引等于白建立。

覆盖索引，尽量不要写星号，这种低效率的查询。

 
select * from lyshark where id=1;     -- 完蛋的。
select x,y,z from lyshark wehre id =1;   -- 覆盖到主键上，查询效率提升很多。

using where -> 使用了where using join buffer -> 用到了缓存buffer

实现单表索引优化

 
create table if not exists article(
id int(10) unsigned not null primary key auto_increment,
author_id int(10) unsigned not null,
category_id int(10) unsigned not null,
views int(10) unsigned not null,
comments int(10) unsigned not null,
title varbinary(255) not null,
content text not null );
 
insert into article(author_id,category_id,views,comments,title,content) values(1,1,1,1,'1','1'),(1,1,1,2,'1','1'),(2,2,2,2,'2','2'),(3,3,3,3,'3','3');
 
 
MariaDB [lyshark]> explain select id,author_id from article where category_id=1 and comments>1 order by views desc limit 1;
 
-- 创建复合索引 all 变为了range 只解决了全表扫描问题
MariaDB [lyshark]> create index idx_article_ccv on article(category_id,comments,views);
 
-- 最后一个完整版的
MariaDB [lyshark]> drop index idx_article_ccv on article;
MariaDB [lyshark]> create index idx_article_cv on article(category_id,views);

实现两表索引优化

 
create table class
(
    id int(10) auto_increment,
    card int(10) not null,
    primary key(id)
);
create table book
(
    bookid int(10) auto_increment,
    card int(10) not null,
    primary key(bookid)
);
 
MariaDB [lyshark]> insert into class(card) values(floor(1+(rand()*20)));  * 10
MariaDB [lyshark]> insert into book(card) values(floor(1+(rand()*20)));  * 10
MariaDB [lyshark]> select * from book inner join class on book.card = class.card;
 
-- 左连接的特性是左表全都有,连接右表的部分
-- 左右链接总有一张表是用来驱动的,左连接链接的是右表,如下左表class右表是book
MariaDB [lyshark]> select * from class left join book on book.card = class.card;
 
-- 左连接情况下,将索引建立在右表上面效率是最高的,如下右表是book
MariaDB [lyshark]> alter table book add index left_index(card);
 
-- 右链接,需要将索引加到左边表上,也就是加到class表的card字段上.
MariaDB [lyshark]> show index from book;
MariaDB [lyshark]> drop index left_index on book;
 
MariaDB [lyshark]> select * from class right join book on book.card = class.card;
MariaDB [lyshark]> alter table class add index right_index(card);

三张表索引优化

 
create table phone
(
    phoneid int(10) auto_increment,
    card int(10) not null,
    primary key(phoneid)
);
 
MariaDB [lyshark]> insert into phone(card) values(floor(1+(rand()*20)));  * 10
 
 
-- 最简单的链接查询
MariaDB [lyshark]> select * from class inner join book on class.card=book.card inner join phone on book.card=phone.card;
 
 
-- 左连接查询
MariaDB [lyshark]> explain select * from class left join book on class.card=book.card left join phone on book.card=phone.card;
 
alter table book add index book_left_index(card);
alter table phone add index phone_left_index(card);
 
-- join 语句优化建议
 
-- 1. 尽可能减少join语句中的NestedLoop的循环次数: 永远用小结果集,驱动大的结果集.
-- 2. 优先优化NestedLoop的内层循环
-- 保证join语句中被驱动表上join条件字段已经被索引.
 
-- 当无法保证被驱动表的join条件字段被索引且内存资源充足的前提下,不要太吝惜joinbuffer 的设置.
-- 如果是三表,左查询,那么我们应该将索引,建立在左连接表中.
 
 
-- 右连接查询
MariaDB [lyshark]> explain select * from class right join book on class.card=book.card right join phone on book.card=phone.card;
 
alter table book add index book_left_index(card);
alter table phone add index phone_left_index(card);

解决中文乱码问题:

 
[root@localhost mysql]# cp -a /usr/share/mysql/my-huge.cnf /etc/my.cnf
[root@localhost mysql]# vim /etc/my.cnf
[client]
default-character-set=utf8
 
[mysqld]
character_set_server=utf8
character_set_client=utf8
collation-server=utf8_general_ci
 
[mysql]
default-character-set=utf8
 
-- 查询字符集编码
MariaDB [lyshark]> select * from information_schema.character_sets;
MariaDB [lyshark]> show character set like 'utf8%';
MariaDB [lyshark]> show variables like 'character_set%';
 
-- 设置全局字符集
set global character_set_client=utf8;
set global character_set_connection=utf8;
set global character_set_database=utf8;
set global character_set_results=utf8;
set global character_set_server=utf8;
 
-- 更新指定表为utf8格式
MariaDB [lyshark]> alter database lyshark default character set utf8 collate utf8_general_ci;
MariaDB [lyshark]> alter table lyshark.user convert to character set utf8 collate utf8_general_ci;
-- 错误日志
[root@localhost ~]# vim /etc/my.cnf
 
[mysqld]
log-error="/var/log/mariadb/mariadb.log"
 
MariaDB [(none)]> show variables like 'log%';
[root@localhost ~]# cat /var/log/mariadb/mariadb.log |head -n 10
 
MariaDB [(none)]> flush logs;
 
 
-- 二进制日志
[root@localhost ~]# vim /etc/my.cnf
 
[mysqld]
log-bin="/tmp"              #设置开启日志,也可不指定日志保存位置
expire_logs_days = 10       #设置日志自动清理天数
max_binlog_size = 100M      #定义了单个文件的大小限制
 
 
-- 删除日志
MariaDB [(none)]> show binary logs;
MariaDB [(none)]> purge master logs to "mariadb-bin.000001";
MariaDB [(none)]> purge master logs before "20180101";
 
[root@localhost ~]# mysqlbinlog mariadb-bin.000001
 
 
 
 
 
-- 慢查询日志
MariaDB [lyshark]> show variables like '%slow_query_log%';
MariaDB [lyshark]> set global slow_query_log=1;
 
MariaDB [lyshark]> show variables like '%long_query_time%';
MariaDB [lyshark]> set global long_query_time=3;
MariaDB [lyshark]> show global status like '%Slow_queries%';
 
[root@localhost mysql]# cat /var/lib/mysql/localhost-slow.log
[root@localhost ~]# vim /etc/my.cnf
 
[mysqld]
log-slow-queries="/var/lib/mysql/localhost-slow.log"
long_query_time=10
log_output=FILE
 
 
 
-s 排序方式
-c 访问次数
-l 锁定时间
-r 返回记录
-t 查询时间
-al 平均锁定时间
-ar 平均返回记录数
-at 平均查询时间
-t 返回前面多少条记录
-g 匹配正则
 
-- 得到返回记录集最多的10个SQL
[root@localhost mysql]# mysqldumpslow -s -r -t 10 /var/lib/mysql/localhost-slow.log
 
-- 得到访问次数最多的10个SQL
[root@localhost mysql]# mysqldumpslow -s -c -t 10 /var/lib/mysql/localhost-slow.log
 
-- 得到按照时间排序的前十条里面含有左连接的查询语句.
[root@localhost mysql]# mysqldumpslow -s -t -t 10 -g 'left join' /var/lib/mysql/localhost-slow.log
 
show variables like 'profiling';
set profiling=on;
 
MariaDB [lyshark]> select * from tbl_emp;
MariaDB [lyshark]> show profiles;   // 查询系统中执行的sql
-- 查询3号记录中的问题，得到3号语句的查询生命周期。
MariaDB [lyshark]> show profile cpu,block io for query 3;

	MariaDB [lyshark]> create table book
	-> (
	-> u_id int not null,
	-> u_book varchar(20) not null,
	-> index(u_id)
	-> );

	-- 使用show index语句查看指定表中创建的索引
	MariaDB [lyshark]> show index from book;
	MariaDB [lyshark]> show create table book \G;
	MariaDB [lyshark]> explain select * from book where u_id=1 \G;

	MariaDB [lyshark]> create table table_1
	-> (
	-> id int not null,
	-> name char(30) not null,
	-> unique index UniqIdx(id)
	-> );
	Query OK, 0 rows affected (0.02 sec)

	MariaDB [lyshark]> show create table table_1 \G;
	************************* 1. row *************************
	Table: table_1
	Create Table: CREATE TABLE `table_1` (
	`id` int(11) NOT NULL,
	`name` char(30) NOT NULL,
	UNIQUE KEY `UniqIdx` (`id`) #id字段已经成功建立了一个名为UniqIdx的唯一索引
	) ENGINE=InnoDB DEFAULT CHARSET=latin1
	1 row in set (0.00 sec)

	MariaDB [lyshark]> create table table_2
	-> (
	-> id int not null,
	-> name char(50) null,
	-> index SingleIdx(name(20))
	-> );
	Query OK, 0 rows affected (0.03 sec)

	MariaDB [lyshark]> show create table table_2 \G;
	************************* 1. row *************************
	Table: table_2
	Create Table: CREATE TABLE `table_2` (
	`id` int(11) NOT NULL,
	`name` char(50) DEFAULT NULL,
	KEY `SingleIdx` (`name`(20)) #name字段上已经成功建立了一个单列索引,名称为SingleIdx
	) ENGINE=InnoDB DEFAULT CHARSET=latin1
	1 row in set (0.00 sec)

	MariaDB [lyshark]> create table table_3
	-> (
	-> id int not null,
	-> name char(30) not null,
	-> age int not null,
	-> info varchar(255),
	-> index MultiIdx(id,name,age)
	-> );

	-- 注意：MySQL5.7默认存储引擎室是InnoDB,在这里我们要改成MyISAM,不然索引会报错
	MariaDB [lyshark]> create table table_4(
	-> id int not null,
	-> name char(40) not null,
	-> age int not null,
	-> info varchar(255),
	-> fulltext index FullTxtIdx(info)
	-> )engine=MyISAM;

	MariaDB [lyshark]> create table table_5
	-> (
	-> g geometry not null,
	-> spatial index spatIdx(g)
	-> )engine=MyISAM;

	MariaDB [lyshark]> create table book
	-> (
	-> bookid int not null,
	-> bookname varchar(255) not null,
	-> authors varchar(255) not null,
	-> info varchar(255) null,
	-> comment varchar(255) null,
	-> year_public year not null
	-> );

	-- 添加普通索引
	MariaDB [lyshark]> alter table book add index BKNameIdx(bookname(30));

	-- 添加唯一索引
	MariaDB [lyshark]> alter table book add unique index UniqidIdx(bookId);

	-- 添加单列索引
	MariaDB [lyshark]> alter table book add index BkcmtIdx(comment(50));

	-- 添加组合索引
	MariaDB [lyshark]> alter table book add index BkAuAndInfoIdx(authors(30),info(50));

	-- 通过索引名字删除索引
	MariaDB [lyshark]> alter table book drop index UniqidIdx;
	MariaDB [lyshark]> alter table book drop index BKNameIdx;

	MariaDB [lyshark]> explain select s_name,s_city from suppliers where s_id IN (select Gid from lyshark where Uid='a1');
	+------+-------------+-----------+-------+---------------+---------+---------+-------+------+-------+
	\| id \| select_type \| table \| type \| possible_keys \| key \| key_len \| ref \| rows \| Extra \|
	+------+-------------+-----------+-------+---------------+---------+---------+-------+------+-------+
	\| 1 \| PRIMARY \| lyshark \| const \| PRIMARY \| PRIMARY \| 30 \| const \| 1 \| \|
	\| 1 \| PRIMARY \| suppliers \| const \| PRIMARY \| PRIMARY \| 4 \| const \| 1 \| \|
	+------+-------------+-----------+-------+---------------+---------+---------+-------+------+-------+

	MariaDB [lyshark]> explain select Course.CID,Course.Cname from Course join(
	select CID from lyshark.StudentScore where SID = (select SID from lyshark.Student where Sname='周梅')
	)as StudentScore on Course.CID = StudentScore.CID;
	+---------+-------------+--------------+------+---------------+------+---------+------+------+----------+
	\| id \| select_type \| table \| type \| possible_keys \| key \| key_len \| ref \| rows \| Extra \|
	+---------+-------------+--------------+------+---------------+------+---------+------+------+----------+
	\| 1(1) \| SIMPLE \| Course \| ALL \| NULL \| NULL \| NULL \| NULL \| 3 \| \|
	\| 1(2) \| SIMPLE \| StudentScore \| ALL \| NULL \| NULL \| NULL \| NULL \| 18 \| \|
	\| 3 \| SUBQUERY \| Student \| ALL \| NULL \| NULL \| NULL \| NULL \| 10 \| \|
	+---------+-------------+--------------+------+---------------+------+---------+------+------+----------+

	ID字段的理解:
	1.当ID字段相同的情况下执行数据是从上到下,例如第一张表中会由上至下执行下来.
	2.当ID不同的情况,如果是子查询,id越大的将在最前面被执行,例如第二张表执行顺序为3->1(1)->1(2)

	MariaDB [lyshark]> explain select * from tbl_emp a left join tbl_dept b on a.deptld=b.id where b.id is null
	-> union
	-> select * from tbl_emp a right join tbl_dept b on a.deptld = b.id where a.deptld is null;
	+------+--------------+------------+--------+---------------+---------+---------+------------------+------+---------+
	\| id \| select_type \| table \| type \| possible_keys \| key \| key_len \| ref \| rows \| Extra \|
	+------+--------------+------------+--------+---------------+---------+---------+------------------+------+---------+
	\| 1 \| PRIMARY \| a \| ALL \| NULL \| NULL \| NULL \| NULL \| 8 \| \|
	\| 1 \| PRIMARY \| b \| eq_ref \| PRIMARY \| PRIMARY \| 4 \| lyshark.a.deptld \| 1 \| \|
	\| 2 \| UNION \| b \| ALL \| NULL \| NULL \| NULL \| NULL \| 5 \| \|
	\| 2 \| UNION \| a \| ALL \| fk_dept_id \| NULL \| NULL \| NULL \| 8 \| \|
	\| NULL \| UNION RESULT \| <union1,2> \| ALL \| NULL \| NULL \| NULL \| NULL \| NULL \| \|
	+------+--------------+------------+--------+---------------+---------+---------+------------------+------+---------+

	SIMPLE => 标志着简单的select查询请求,查询中不包含子查询或者union查询.
	PRIMARY => 查询中任何复杂的查询中,最外层的查询语句,就是最后加载的语句.
	SUBQUERY => 子查询类型,在select或where列表中包含了子查询.
	DERIVED => 在FROM列表中包含子查询,会被标记为DERIVED(衍生),此时会递归执行子查询,并存储在临时表中.
	UNION => 若第二个SELECT出现在UNION之后,则标记为UNION.
	UNION RESULT => 从UNION表中获取结果的SELECT

	system -> const -> eq_ref -> ref -> range ->index -> all


	system -> 表中只有一条记录，这是const类型的特里，平时不会出现。

	const -> 主键唯一索引：表示通过索引一次就找到数据，例如查询一个常量。
	MariaDB [lyshark]> explain select * from lyshark where Uid="a1";
	+------+-------------+---------+-------+---------------+---------+---------+-------+------+-------+
	\| id \| select_type \| table \| type \| possible_keys \| key \| key_len \| ref \| rows \| Extra \|
	+------+-------------+---------+-------+---------------+---------+---------+-------+------+-------+
	\| 1 \| SIMPLE \| lyshark \| const \| PRIMARY \| PRIMARY \| 30 \| const \| 1 \| \|
	+------+-------------+---------+-------+---------------+---------+---------+-------+------+-------+


	eq_ref -> 唯一性索引扫描，对于每个索引键，表中只有一条记录与之匹配，常用于主键或唯一索引扫描。

	ref-> 非唯一性索引扫描，返回匹配某个单独值的所有行，被之上也是一种索引访问。查找扫描混合体

	MariaDB [lyshark]> create table t1(col1 int,col2 int);
	MariaDB [lyshark]> alter table t1 add index idx_col1_col2(col1,col2);
	MariaDB [lyshark]> explain select * from t1 where col1=1;
	+------+-------------+-------+------+---------------+---------------+---------+-------+------+--------+
	\| id \| select_type \| table \| type \| possible_keys \| key \| key_len \| ref \| rows \| Extra \|
	+------+-------------+-------+------+---------------+---------------+---------+-------+------+--------+
	\| 1 \| SIMPLE \| t1 \| ref \| idx_col1_col2 \| idx_col1_col2 \| 5 \| const \| 1 \| \|
	+------+-------------+-------+------+---------------+---------------+---------+-------+------+--------+

	range -> 范围扫描，只检索给定范围的行，key列显示使用了那个索引。 where，between，<>，in 等查询中使用。

	explain select * from t1 where col1 between 1 and 2;
	explain select * from t1 where col1 in (1,2,3);

	index -> 全索引扫描，全表索引扫描，比all要好一些。
	MariaDB [lyshark]> explain select * from t1;

	all -> 全表扫描。最差的性能。

	possible_keys => 显示可能应用在这张表中的索引，一个或多个，该索引会被列出，但不一定被实际查询使用。
	key => 实际应用到本次查询的索引类型。最重要的。如果为NULL，则说明没有使用索引。


	MariaDB [lyshark]> explain select col1,col2 from t1;
	+------+-------------+-------+-------+---------------+---------------+---------+------+------+-------+
	\| id \| select_type \| table \| type \| possible_keys \| key \| key_len \| ref \| rows \| Extra \|
	+------+-------------+-------+-------+---------------+---------------+---------+------+------+-------+
	\| 1 \| SIMPLE \| t1 \| index \| NULL \| idx_col1_col2 \| 10 \| NULL \| 1 \| \|
	+------+-------------+-------+-------+---------------+---------------+---------+------+------+-------+


	------------------------------------------------------------------------------
	查询中若使用了覆盖索引，则该索引仅出现key列表中覆盖索引以下就是原理

	MariaDB [lyshark]> create table t1(col1 int,col2 int,col3 int);
	MariaDB [lyshark]> alter table t1 add index idx_col1_col2(col1,col2);

	-- 建立的索引与，查询的行数，必须一致，col1,col2是有索引的。
	MariaDB [lyshark]> explain select col1,col2 from t1;
	+------+-------------+-------+-------+---------------+---------------+---------+------+------+-------------+
	\| id \| select_type \| table \| type \| possible_keys \| key \| key_len \| ref \| rows \| Extra \|
	+------+-------------+-------+-------+---------------+---------------+---------+------+------+-------------+
	\| 1 \| SIMPLE \| t1 \| index \| NULL \| idx_col1_col2 \| 10 \| NULL \| 1 \| Using \|
	+------+-------------+-------+-------+---------------+---------------+---------+------+------+-------------+

	-- 扫描三个值，不会出现使用索引的情况。
	MariaDB [lyshark]> explain select col1,col2,col3 from t1;
	+------+-------------+-------+------+---------------+------+---------+------+------+-------+
	\| id \| select_type \| table \| type \| possible_keys \| key \| key_len \| ref \| rows \| Extra \|
	+------+-------------+-------+------+---------------+------+---------+------+------+-------+
	\| 1 \| SIMPLE \| t1 \| ALL \| NULL \| NULL \| NULL \| NULL \| 1 \| \|
	+------+-------------+-------+------+---------------+------+---------+------+------+-------+

	key_len 表示索引中使用的字节数，这个长度用的越少越好，kenLen长度是根据表的定义计算得出，而不是表中数据检索出的。




	ref 显示索引的那一列被使用了，如果可能的话，是一个常数，那些列或常量被用于查找索引列上的值。定义了引用了那些库。

	rows 根据表统计信息及索引选用情况，大致估算出找到所需的记录所需要读取的行数，该值当然也是越小越好。每张表有多少行被优化器查询。

	MariaDB [lyshark]> explain select * from lyshark;
	+------+-------------+---------+------+---------------+------+---------+------+------+-------+
	\| id \| select_type \| table \| type \| possible_keys \| key \| key_len \| ref \| rows \| Extra \|
	+------+-------------+---------+------+---------------+------+---------+------+------+-------+
	\| 1 \| SIMPLE \| lyshark \| ALL \| NULL \| NULL \| NULL \| NULL \| 17 \| \|
	+------+-------------+---------+------+---------------+------+---------+------+------+-------+
	1 row in set (0.00 sec)


	MariaDB [lyshark]> create table tt1(id int primary key,col1 varchar(10),col2 varchar(10));
	MariaDB [lyshark]> create table tt2(id int primary key,col1 varchar(10),col2 varchar(10));
	MariaDB [lyshark]> create index idt_col1_col2 on tt2(col1,col2);

	MariaDB [lyshark]> explain select * from tt1,tt2 where tt1.id = tt2.id and tt2.col1 = 'admin';
	+------+-------------+-------+--------+-----------------------+---------+---------+----------------+------+-------------+
	\| id \| select_type \| table \| type \| possible_keys \| key \| key_len \| ref \| rows \| Extra \|
	+------+-------------+-------+--------+-----------------------+---------+---------+----------------+------+-------------+
	\| 1 \| SIMPLE \| tt1 \| ALL \| PRIMARY \| NULL \| NULL \| NULL \| 1 \| \|
	\| 1 \| SIMPLE \| tt2 \| eq_ref \| PRIMARY,idt_col1_col2 \| PRIMARY \| 4 \| lyshark.tt1.id \| 1 \| Using where \|
	+------+-------------+-------+--------+-----------------------+---------+---------+----------------+------+-------------+
	2 rows in set (0.00 sec)

	MariaDB [lyshark]> create table tab1(id int primary key,col1 int,col2 int,col3 int);
	MariaDB [lyshark]> insert into tab1 values(1,1,2,3),(2,4,5,6),(3,7,8,9);
	MariaDB [lyshark]> create index tab1_col1_col2_col3 on tab1(col1,col2,col3);

	MariaDB [lyshark]> explain select col1 from tab1 where col1 order by col3 \G
	************************* 1. row *************************
	id: 1
	select_type: SIMPLE
	table: tab1
	type: index
	possible_keys: tab1_col1_col2_col3
	key: tab1_col1_col2_col3
	key_len: 15
	ref: NULL
	rows: 3
	Extra: Using where; Using index; Using filesort
	1 row in set (0.00 sec)

	MariaDB [lyshark]> explain select col1 from tab1 where col1 order by col2,col3 \G

	查询使用索引，没问题的，只是说，我们没有针对order by 建立排序索引，或者是建立了索引，你没用上！！


	以下我们加上全部索引字段，从此下面这条sql性能更高了。
	MariaDB [lyshark]> explain select col1 from tab1 where col1 order by col1,col2,col3 \G
	************************* 1. row *************************
	id: 1
	select_type: SIMPLE
	table: tab1
	type: index
	possible_keys: tab1_col1_col2_col3
	key: tab1_col1_col2_col3
	key_len: 15
	ref: NULL
	rows: 3
	Extra: Using where; Using index
	1 row in set (0.00 sec)


	如果可以，尽快优化。

	select * from lyshark where id=1; -- 完蛋的。
	select x,y,z from lyshark wehre id =1; -- 覆盖到主键上，查询效率提升很多。

	create table if not exists article(
	id int(10) unsigned not null primary key auto_increment,
	author_id int(10) unsigned not null,
	category_id int(10) unsigned not null,
	views int(10) unsigned not null,
	comments int(10) unsigned not null,
	title varbinary(255) not null,
	content text not null );

	insert into article(author_id,category_id,views,comments,title,content) values(1,1,1,1,'1','1'),(1,1,1,2,'1','1'),(2,2,2,2,'2','2'),(3,3,3,3,'3','3');


	MariaDB [lyshark]> explain select id,author_id from article where category_id=1 and comments>1 order by views desc limit 1;

	-- 创建复合索引 all 变为了range 只解决了全表扫描问题
	MariaDB [lyshark]> create index idx_article_ccv on article(category_id,comments,views);

	-- 最后一个完整版的
	MariaDB [lyshark]> drop index idx_article_ccv on article;
	MariaDB [lyshark]> create index idx_article_cv on article(category_id,views);

	create table class
	(
	id int(10) auto_increment,
	card int(10) not null,
	primary key(id)
	);
	create table book
	(
	bookid int(10) auto_increment,
	card int(10) not null,
	primary key(bookid)
	);

	MariaDB [lyshark]> insert into class(card) values(floor(1+(rand()20))); 10
	MariaDB [lyshark]> insert into book(card) values(floor(1+(rand()20))); 10
	MariaDB [lyshark]> select * from book inner join class on book.card = class.card;

	-- 左连接的特性是左表全都有,连接右表的部分
	-- 左右链接总有一张表是用来驱动的,左连接链接的是右表,如下左表class右表是book
	MariaDB [lyshark]> select * from class left join book on book.card = class.card;

	-- 左连接情况下,将索引建立在右表上面效率是最高的,如下右表是book
	MariaDB [lyshark]> alter table book add index left_index(card);

	-- 右链接,需要将索引加到左边表上,也就是加到class表的card字段上.
	MariaDB [lyshark]> show index from book;
	MariaDB [lyshark]> drop index left_index on book;

	MariaDB [lyshark]> select * from class right join book on book.card = class.card;
	MariaDB [lyshark]> alter table class add index right_index(card);

	create table phone
	(
	phoneid int(10) auto_increment,
	card int(10) not null,
	primary key(phoneid)
	);

	MariaDB [lyshark]> insert into phone(card) values(floor(1+(rand()20))); 10


	-- 最简单的链接查询
	MariaDB [lyshark]> select * from class inner join book on class.card=book.card inner join phone on book.card=phone.card;


	-- 左连接查询
	MariaDB [lyshark]> explain select * from class left join book on class.card=book.card left join phone on book.card=phone.card;

	alter table book add index book_left_index(card);
	alter table phone add index phone_left_index(card);

	-- join 语句优化建议

	-- 1. 尽可能减少join语句中的NestedLoop的循环次数: 永远用小结果集,驱动大的结果集.
	-- 2. 优先优化NestedLoop的内层循环
	-- 保证join语句中被驱动表上join条件字段已经被索引.

	-- 当无法保证被驱动表的join条件字段被索引且内存资源充足的前提下,不要太吝惜joinbuffer 的设置.
	-- 如果是三表,左查询,那么我们应该将索引,建立在左连接表中.


	-- 右连接查询
	MariaDB [lyshark]> explain select * from class right join book on class.card=book.card right join phone on book.card=phone.card;

	alter table book add index book_left_index(card);
	alter table phone add index phone_left_index(card);

	[root@localhost mysql]# cp -a /usr/share/mysql/my-huge.cnf /etc/my.cnf
	[root@localhost mysql]# vim /etc/my.cnf
	[client]
	default-character-set=utf8

	[mysqld]
	character_set_server=utf8
	character_set_client=utf8
	collation-server=utf8_general_ci

	[mysql]
	default-character-set=utf8

	-- 查询字符集编码
	MariaDB [lyshark]> select * from information_schema.character_sets;
	MariaDB [lyshark]> show character set like 'utf8%';
	MariaDB [lyshark]> show variables like 'character_set%';

	-- 设置全局字符集
	set global character_set_client=utf8;
	set global character_set_connection=utf8;
	set global character_set_database=utf8;
	set global character_set_results=utf8;
	set global character_set_server=utf8;

	-- 更新指定表为utf8格式
	MariaDB [lyshark]> alter database lyshark default character set utf8 collate utf8_general_ci;
	MariaDB [lyshark]> alter table lyshark.user convert to character set utf8 collate utf8_general_ci;
	-- 错误日志
	[root@localhost ~]# vim /etc/my.cnf

	[mysqld]
	log-error="/var/log/mariadb/mariadb.log"

	MariaDB [(none)]> show variables like 'log%';
	[root@localhost ~]# cat /var/log/mariadb/mariadb.log \|head -n 10

	MariaDB [(none)]> flush logs;


	-- 二进制日志
	[root@localhost ~]# vim /etc/my.cnf

	[mysqld]
	log-bin="/tmp" #设置开启日志,也可不指定日志保存位置
	expire_logs_days = 10 #设置日志自动清理天数
	max_binlog_size = 100M #定义了单个文件的大小限制


	-- 删除日志
	MariaDB [(none)]> show binary logs;
	MariaDB [(none)]> purge master logs to "mariadb-bin.000001";
	MariaDB [(none)]> purge master logs before "20180101";

	[root@localhost ~]# mysqlbinlog mariadb-bin.000001





	-- 慢查询日志
	MariaDB [lyshark]> show variables like '%slow_query_log%';
	MariaDB [lyshark]> set global slow_query_log=1;

	MariaDB [lyshark]> show variables like '%long_query_time%';
	MariaDB [lyshark]> set global long_query_time=3;
	MariaDB [lyshark]> show global status like '%Slow_queries%';

	[root@localhost mysql]# cat /var/lib/mysql/localhost-slow.log
	[root@localhost ~]# vim /etc/my.cnf

	[mysqld]
	log-slow-queries="/var/lib/mysql/localhost-slow.log"
	long_query_time=10
	log_output=FILE



	-s 排序方式
	-c 访问次数
	-l 锁定时间
	-r 返回记录
	-t 查询时间
	-al 平均锁定时间
	-ar 平均返回记录数
	-at 平均查询时间
	-t 返回前面多少条记录
	-g 匹配正则

	-- 得到返回记录集最多的10个SQL
	[root@localhost mysql]# mysqldumpslow -s -r -t 10 /var/lib/mysql/localhost-slow.log

	-- 得到访问次数最多的10个SQL
	[root@localhost mysql]# mysqldumpslow -s -c -t 10 /var/lib/mysql/localhost-slow.log

	-- 得到按照时间排序的前十条里面含有左连接的查询语句.
	[root@localhost mysql]# mysqldumpslow -s -t -t 10 -g 'left join' /var/lib/mysql/localhost-slow.log

	show variables like 'profiling';
	set profiling=on;

	MariaDB [lyshark]> select * from tbl_emp;
	MariaDB [lyshark]> show profiles; // 查询系统中执行的sql
	-- 查询3号记录中的问题，得到3号语句的查询生命周期。
	MariaDB [lyshark]> show profile cpu,block io for query 3;