select … into outfile 'path' 备份
此种方式恢复速度非常快，比insert的插入速度要快的多，他跟有备份功能丰富的mysqldump不同的是，他只能备份表中的数据，并不能包含表的结构。如果备份完成之后，表被drop,是无法实现恢复操作的（除非有表结构）。
mysql> select * from t1 into outfile '/mydata/mysql/mysql3307/data/t1.sql';
ERROR 1290 (HY000): The MySQL server is running with the --secure-file-priv option so it cannot execute this statement
mysql> show variables like '%secure%';
+--------------------------+-------+
| Variable_name | Value |
+--------------------------+-------+
| require_secure_transport | OFF |
| secure_auth | ON |
| secure_file_priv | NULL |
+--------------------------+-------+
3 rows in set (0.00 sec)
mysql> set secure_file_priv='/tmp';
ERROR 1238 (HY000): Variable 'secure_file_priv' is a read only variable
mysql> set session secure_file_priv=='/tmp';
ERROR 1064 (42000): You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near '=='/tmp'' at line 1
mysql>
关闭MYSQL数据库，设置secure_file_priv 目录。
[root@mysql5 ~]# mysqladmin -S /tmp/mysql3307.sock -uroot -pmysql shutdown
mysqladmin: [Warning] Using a password on the command line interface can be insecure.
[root@mysql5 ~]#
[root@mysql5 ~]#
[root@mysql5 ~]# ps -ef |grep mysql
root 3506 2071 0 01:24 pts/1 00:00:00 grep mysql
[root@mysql5 ~]#
vi /etc/my3307.cnf
[mysqld]里面加入
secure_file_priv=/tmp
启动后查看
mysql> show global variables like '%secu%';
+--------------------------+-------+
| Variable_name | Value |
+--------------------------+-------+
| require_secure_transport | OFF |
| secure_auth | ON |
| secure_file_priv | /tmp/ |
+--------------------------+-------+
3 rows in set (0.01 sec)
select * from t1 into outfile '/tmp/t1_onlydata.sql';
mysql> select * from t1 into outfile '/tmp/t1_onlydata.sql';
Query OK, 972864 rows affected (1.12 sec)
mysql> desc t1;
+-------+--------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+-------+--------+------+-----+---------+-------+
| id | int(4) | YES | MUL | NULL | |
+-------+--------+------+-----+---------+-------+
1 row in set (0.00 sec)
全是文本文件数据。
select … into outfile 'path' 恢复
清除t1表数据，并进行恢复
mysql> truncate table t1;
Query OK, 0 rows affected (0.04 sec)
mysql> desc t1;
+-------+--------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+-------+--------+------+-----+---------+-------+
| id | int(4) | YES | MUL | NULL | |
+-------+--------+------+-----+---------+-------+
1 row in set (0.00 sec)
LOAD DATA恢复数据（只能恢复数据）
语法：
LOAD DATA [LOW_PRIORITY | CONCURRENT] [LOCAL] INFILE 'file_name'
[REPLACE | IGNORE]
INTO TABLE tbl_name
[PARTITION (partition_name [, partition_name] ...)]
[CHARACTER SET charset_name]
[{FIELDS | COLUMNS}
[TERMINATED BY 'string']
[[OPTIONALLY] ENCLOSED BY 'char']
[ESCAPED BY 'char']
]
[LINES
[STARTING BY 'string']
[TERMINATED BY 'string']
]
[IGNORE number {LINES | ROWS}]
[(col_name_or_user_var
[, col_name_or_user_var] ...)]
[SET col_name={expr | DEFAULT},
[, col_name={expr | DEFAULT}] ...]
load data INFILE '/tmp/t1_onlydata.sql' INTO TABLE test.t1;
查看表数据：
恢复成功。
load data 与insert的插入速度对比
于是，我猜想可以用mysqldump进行表结构的备份，用select …into outfile 备份数据，load data 恢复数据，测试两者速度。
创建一张表。
delimiter //
create procedure per2()
begin
declare i int;
set i=1;
while i <= 1000000 do
insert into test.L values(i,'aaaaa');
set i=i+1;
end while;
end
//
执行存储过程：
call per2();
//
查看数据（实际我只插入仅14万行数据）
mysqldump备份元数据文件
mysqldump -S /tmp/mysql3307.sock -uroot -pmysql --single-transaction --set-gtid-purged=OFF test L -d > /tmp/L_meta.sql
select … into outfile 'path' 备份
select * from test.L into outfile '/tmp/20180525test_Ldata.sql'
mysqldump备份整个表
mysqldump -S /tmp/mysql3307.sock -uroot -pmysql --single-transaction --set-gtid-purged=OFF test L > /tmp/L_table.sql
恢复时间对比
1、MYSQLdump先恢复表结构（或者直接建表）
先删除，恢复表结构。
恢复表结构：
这核时间很快，就是建一张表的时间，不计算进去对整体时间没有影响。
例如我建表：
建表时间0.03s。
2、load data恢复时间
load data INFILE '/tmp/20180525test_Ldata.sql' INTO TABLE test.L;
时间是1.59s。
3、MYSQLDUMP这种插入方式恢复
time mysql -S /tmp/mysql3307.sock -uroot -pmysql test < /tmp/L_table.sql
将近2s MYSQLDUMP恢复时间。
总结
元数据恢复+LOAD DATA时间一共
T1=0.03+1.59=1.62s
MYSQLDUMP恢复一共花了
T2=1.99s
对于14万数据，load data比mysqldump快了近1/4，对于大数据量，应该快更多，在允许的情况下，可以利用元数据（或者表结构），配合LOAD DATA恢复单表。