标签:dev mil mysqld ls -l oracle数据库 back in place evel under
在MySQL数据库中,没有类似于SQL Server数据库或Oracle数据库中索引重建的语法(ALTER INDEX ... REBUILD),那么在MySQL数据库中,是否有什么方式重建索引呢? 在官方文档中"2.11.10 Rebuilding or Repairing Tables or Indexes"中,提到下面三种方式可以Rebuild Index
· Dump and Reload Method
· ALTER TABLE Method
· REPAIR TABLE Method
另外, OPTIMIZE TABLE也会对索引进行重建,下面我们来简单验证、测试一下,如有不对或不足的地方,敬请指正。
第一种方法(mysqldump导出然后重新导入),相当于重新CREATE INDEXES , 这里就不讨论了。下面我们来看看其它几种方法,那么要判断索引是否REBUILD了呢?我们来测试验证一下吧,新建测试表如下:
CREATE TABLE t1 (
c1 INT NOT NULL AUTO_INCREMENT PRIMARY KEY,
c2 VARCHAR(100),
c3 VARCHAR(100) )
ENGINE=InnoDB;
create index ix_t1_c2 on t1(c2);
DROP INDEX + CREATE INDEX方法
这种方法过于简单,这里不叙说了。其实也没有啥好说的。
ALTER TABLE方法
那么我们能否在MySQL中找到索引的创建或修改时间呢?经过查证,目前而言,MySQL中是没有相关系统表或视图会记录索引的创建时间的,我们可以用间接的方法来间接验证,有些方法不是特别可靠和准确,最准确的方法应该是阅读源码:
1:表的创建时间,可以间接推断索引的创建时间。因为索引的创建时间肯定在表的创建时间之后。
2:对应表的idb文件的修改或创建时间(若文件从创建后不曾修改过则可认为创建时间=修改时间,关于更多详细内容,参考“Linux如何查找文件的创建时间”),当然这种方法不是非常准确。我们知道,对于InnoDB存储引擎的表而言,对应的索引数据存储在ibd文件中,所以文件的创建时间或修改时间是间接判断索引创建时间。如果存储引擎为MyISAM的话,还有专门的索引文件MYI。
注意:show indexes from tablename不会显示索引创建时间
mysql> SELECT table_name,create_time FROM information_schema.TABLES WHERE table_name=‘t1‘;
+------------+---------------------+
| TABLE_NAME | CREATE_TIME |
+------------+---------------------+
| t1 | 2019-10-20 08:18:33 |
+------------+---------------------+
1 row in set (0.01 sec)
然后我们对表进行ALTER TABLE t1 ENGINE = InnoDB;进行操作后,然后去验证表的创建时间,如下所示,其实ALTER TABLE xxx ENGINE=InnoDB 其实等价于REBUILD表(REBUILD表就是重建表的意思),所以索引也等价于重新创建了。
在另外一个窗口,我们对比t1.ibd的创建时间,如下所示,也间接验证了表和索引都REBUILD了。(这里是MySQL 8.0.18 ,如果是之前的版本,还有frm之类的文件。)
[root@db-server MyDB]# ls -lrt t1*
-rw-r-----. 1 mysql mysql 131072 Oct 20 08:18 t1.ibd
[root@db-server MyDB]# stat t1.ibd
File: ‘t1.ibd’
Size: 131072 Blocks: 224 IO Block: 4096 regular file
Device: fd00h/64768d Inode: 106665154 Links: 1
Access: (0640/-rw-r-----) Uid: ( 1000/ mysql) Gid: ( 1000/ mysql)
Context: system_u:object_r:mysqld_db_t:s0
Access: 2019-10-20 08:18:25.911990445 +0800
Modify: 2019-10-20 08:18:33.626989940 +0800
Change: 2019-10-20 08:18:33.626989940 +0800
Birth: -
[root@db-server MyDB]# stat t1.ibd
File: ‘t1.ibd’
Size: 131072 Blocks: 224 IO Block: 4096 regular file
Device: fd00h/64768d Inode: 106665156 Links: 1
Access: (0640/-rw-r-----) Uid: ( 1000/ mysql) Gid: ( 1000/ mysql)
Context: system_u:object_r:mysqld_db_t:s0
Access: 2019-10-20 08:20:50.866980953 +0800
Modify: 2019-10-20 08:20:51.744980896 +0800
Change: 2019-10-20 08:20:51.744980896 +0800
Birth: -
REPAIR TABLE方法
REPAIR TABLE方法用于修复被破坏的表,而且它仅仅能用于MyISAM, ARCHIVE,CSV类型的表。下面的测试环境为MySQL 5.6.41,创建测试表,然后对表进行REPAIR TABLE操作
mysql> CREATE TABLE t (
-> c1 INT NOT NULL AUTO_INCREMENT PRIMARY KEY,
-> c2 VARCHAR(100),
-> c3 VARCHAR(100) )
-> ENGINE=MyISAM;
Query OK, 0 rows affected (0.01 sec)
mysql> SELECT table_name,create_time FROM information_schema.TABLES WHERE table_name=‘t‘;
+------------+---------------------+
| table_name | create_time |
+------------+---------------------+
| t | 2019-10-20 08:35:43 |
+------------+---------------------+
1 row in set (0.00 sec)
然后对表t进行修复操作,发现表的create_time没有变化,如下所示:
mysql> REPAIR TABLE t;
+--------+--------+----------+----------+
| Table | Op | Msg_type | Msg_text |
+--------+--------+----------+----------+
| MyDB.t | repair | status | OK |
+--------+--------+----------+----------+
1 row in set (0.01 sec)
mysql> SELECT table_name,create_time FROM information_schema.TABLES WHERE table_name=‘t‘;
+------------+---------------------+
| table_name | create_time |
+------------+---------------------+
| t | 2019-10-20 08:35:43 |
+------------+---------------------+
1 row in set (0.00 sec)
在另外一个窗口,我们发现索引文件t.MYI的修改时间和状态更改时间都变化了,所以判断索引重建(Index Rebuild)了。
[root@testlnx02 MyDB]# ls -lrt t.*
-rw-rw----. 1 mysql mysql 8608 Oct 20 08:35 t.frm
-rw-rw----. 1 mysql mysql 1024 Oct 20 08:35 t.MYI
-rw-rw----. 1 mysql mysql 0 Oct 20 08:35 t.MYD
[root@testlnx02 MyDB]# stat t.MYI
File: `t.MYI‘
Size: 1024 Blocks: 8 IO Block: 4096 regular file
Device: fd00h/64768d Inode: 1836747 Links: 1
Access: (0660/-rw-rw----) Uid: ( 27/ mysql) Gid: ( 27/ mysql)
Access: 2019-10-20 08:36:02.395428301 +0800
Modify: 2019-10-20 08:35:43.112562600 +0800
Change: 2019-10-20 08:35:43.112562600 +0800
[root@testlnx02 MyDB]# stat t.MYI
File: `t.MYI‘
Size: 1024 Blocks: 8 IO Block: 4096 regular file
Device: fd00h/64768d Inode: 1836747 Links: 1
Access: (0660/-rw-rw----) Uid: ( 27/ mysql) Gid: ( 27/ mysql)
Access: 2019-10-20 08:37:19.686899429 +0800
Modify: 2019-10-20 08:37:10.271475420 +0800
Change: 2019-10-20 08:37:10.271475420 +0800
OPTIMIZE TABLE方法
OPTIMIZE TABLE也可以对索引进行重建,官方文档的介绍如下:
OPTIMIZE TABLE reorganizes the physical storage of table data and associated index data, to reduce storage space and improve I/O efficiency when accessing the table. The exact changes made to each table depend on the storage engine used by that table.
OPTIMIZE TABLE uses online DDL for regular and partitioned InnoDB tables, which reduces downtime for concurrent DML operations. The table rebuild triggered by OPTIMIZE TABLE and performed under the cover by ALTER TABLE ... FORCE is completed in place. An exclusive table lock is only taken briefly during the prepare phase and the commit phase of the operation. During the prepare phase, metadata is updated and an intermediate table is created. During the commit phase, table metadata changes are committed.
OPTIMIZE TABLE rebuilds the table using the table copy method under the following conditions:
·
· When the old_alter_table system variable is enabled.
·
· When the server is started with the --skip-new option.
OPTIMIZE TABLE using online DDL is not supported for InnoDB tables that contain FULLTEXT indexes. The table copy method is used instead.
简单来说,OPTIMIZE TABLE操作使用Online DDL模式修改Innodb普通表和分区表,
该方式会在prepare阶段和commit阶段持有表级锁:在prepare阶段修改表的元数据并且创建一个中间表,在commit阶段提交元数据的修改。
由于prepare阶段和commit阶段在整个事务中的时间比例非常小,可以认为该OPTIMIZE TABLE的过程中不影响表的其他并发操作。
测试验证如下,对表t1做了OPTIMIZE TABLE后, 表的创建时间变成了2019-10-20 08:41:57
mysql> OPTIMIZE TABLE t1;
+---------+----------+----------+-------------------------------------------------------------------+
| Table | Op | Msg_type | Msg_text |
+---------+----------+----------+-------------------------------------------------------------------+
| MyDB.t1 | optimize | note | Table does not support optimize, doing recreate + analyze instead |
| MyDB.t1 | optimize | status | OK |
+---------+----------+----------+-------------------------------------------------------------------+
2 rows in set (0.67 sec)
mysql> SELECT table_name,create_time FROM information_schema.TABLES WHERE table_name=‘t1‘;
+------------+---------------------+
| TABLE_NAME | CREATE_TIME |
+------------+---------------------+
| t1 | 2019-10-20 08:41:57 |
+------------+---------------------+
1 row in set (0.00 sec)
另外,网上有种说法ANALYZE TABLE方法也可以重建索引,其实ANALYZE TABLE是不会对索引进行重建的。测试验证的话,你会发现ibd文件没有变化,表的修改时间/状态更改时间也没有变化。
总结:
测试完后,还是感觉MySQL索引重建的方式怪怪的,可能是有先入为主的观念。总结一下MySQL索引重建的方法:
1: DROP INDEX + RECREATE INDEX.
2: ALTER TABLE方法
3: REPAIR TABLE方法,这种方法对于InnoDB存储引擎的表无效。
4: OPTIMIZE TABLE方法
参考资料:
https://dev.mysql.com/doc/refman/8.0/en/rebuilding-tables.html
https://docs.oracle.com/cd/E17952_01/mysql-5.6-en/rebuilding-tables.html
标签:dev mil mysqld ls -l oracle数据库 back in place evel under
原文地址:https://www.cnblogs.com/kerrycode/p/11706730.html