>데이터 베이스 >MySQL 튜토리얼 >mysql之double write_MySQL

mysql之double write_MySQL

WBOY
WBOY원래의
2016-05-30 17:10:531335검색

1、doublewrite buffer(mysql官方的介绍)

 

InnoDB uses a novel file flush technique called doublewrite. Before writing pages to the data files, InnoDB first writes them to a contiguous area called the doublewrite buffer. Only after the write and the flush to the doublewrite buffer have completed, does InnoDB write the pages to their proper positions in the data file. If there is an operating system, storage subsystem, or mysqld process crash in the middle of a page write, InnoDB can later find a good copy of the page from the doublewrite buffer during crash recovery.

 

Although data is always written twice, the doublewrite buffer does not require twice as much I/O overhead or twice as many I/O operations. Data is written to the buffer itself as a large sequential chunk, with a single fsync() call to the operating system.

 

To turn off the doublewrite buffer, specify the option innodb_doublewrite=0.

 

2、partial page write

 

InnoDB 的Page Size一般是16KB,其数据校验也是针对这16KB来计算的,将数据写入到磁盘是以Page为单位进行操作的。而计算机硬件和操作系统,在极端情况下 (比如断电)往往并不能保证这一操作的原子性,16K的数据,写入4K 时,发生了系统断电/os crash ,只有一部分写是成功的,这种情况下就是 partial page write 问题。

很多DBA 会想到系统恢复后,MySQL 可以根据redo log 进行恢复,而mysql在恢复的过程中是检查page的checksum,checksum就是pgae的最后事务号,发生partial page write 问题时,page已经损坏,找不到该page中的事务号,就无法恢复

 

所以说,当page 损坏之后,其实应用redo是没有意义的,这时候无法使用redo来恢复,因为原始页已经损坏了,会发生数据丢失。

 

3、doublewrite

 

在InnoDB将BP中的Dirty Page刷(flush)到磁盘上时,首先会将(memcpy函数)Page刷到InnoDB tablespace的一个区域中,我们称该区域为Double write Buffer(大小为2MB,每次写入1MB)。在向Double write Buffer写入成功后,第二步、再将数据拷贝到数据文件对应的位置。

 

当第二步过程中发生故障,也就是发生partial page write的问题。恢复的时候先检查页内的checksum是否相同,不一致,则直接从doublewrite中恢复。

 

1)如果写dw buffer失败。那么这些数据不会写到磁盘,innodb会载入磁盘原始数据和redo日志比较,并重新刷到dw buffer。

2)如果写dw buffer成功,但是刷新到磁盘失败,那么innodb就不会通过事务日志来恢复了,而是直接刷新dw buffer中的数据。

 

4、对性能的影响

 

系统需要将数据写两份,一般认为,Double Write是会降低系统性能的。peter猜测可能会有5-10%的性能损失,但是因为实现了数据的一致,是值得的。Mark Callaghan认为这应该是存储层面应该解决的问题,放在数据库层面无疑是牺牲了很多性能的。

 

事实上,Double Write对性能影响并没有你想象(写两遍性能应该降低了50%吧?)的那么大。在BP中一次性往往会有很多的Dirty Page同时被flush,Double Write则把这些写操作,由随机写转化为了顺序写。而在Double Write的第二个阶段,因为Double Write Buffer中积累了很多Dirty Page,所以向真正的数据文件中写数据的时候,可能有很多写操作可以合并,这样有可能会降低Fsync的调用次数。

 

基于上面的原因,Double Write并没有想象的那么糟。另外,Dimitri在测试后,发现打开和关闭Double Write对效率的影响并不大。

 

5、doublewrite参数

 

mysql> show variables like "%double%";
+--------------------+-------+
| Variable_name      | Value |
+--------------------+-------+
| innodb_doublewrite | ON    |
+--------------------+-------+
1 row in set (0.00 sec)

mysql> SHOW STATUS LIKE "%innodb_dblwr%";
+----------------------------+-------+
| Variable_name              | Value |
+----------------------------+-------+
| Innodb_dblwr_pages_written | 447   |
| Innodb_dblwr_writes        | 38    |
+----------------------------+-------+
2 rows in set (0.00 sec)

 

 

从上面可以看出, Flush了447次到doublewrite buffer中,写文件共38次,则每次write合并了447/38次flush

성명:
본 글의 내용은 네티즌들의 자발적인 기여로 작성되었으며, 저작권은 원저작자에게 있습니다. 본 사이트는 이에 상응하는 법적 책임을 지지 않습니다. 표절이나 침해가 의심되는 콘텐츠를 발견한 경우 admin@php.cn으로 문의하세요.