ITPub博客

MySQL InnoDB设置死锁检测的方法

原创 MySQL 作者:chenfeng 时间:2018-05-10 09:31:17 0 删除 编辑
死锁是指两个或两个以上的进程在执行过程中,因争夺资源而造成的一种互相等待的现象,可以认为如果一个资源被锁定,它总会在以后某个时间被释放。而死锁发生在当多个进程访问同一数据库时,其中每个进程拥有的锁都是其他进程所需的,由此造成每个进程都无法继续下去。
InnoDB的并发写操作会触发死锁,InnoDB也提供了死锁检测机制,可以通过设置innodb_deadlock_detect参数可以打开或关闭死锁检测:


innodb_deadlock_detect = on  打开死锁检测,数据库发生死锁时自动回滚(默认选项)
innodb_deadlock_detect = off  关闭死锁检测,发生死锁的时候,用锁超时来处理,通过设置锁超时参数innodb_lock_wait_timeout可以在超时发生时回滚被阻塞的事务


还可以通过设置InnDB Monitors来进一步观察锁冲突详细信息
设置InnoDB Monitors方法
建立test库
mysql>create database test;
Query OK, 1 row affected (0.20 sec)
mysql> use test
Reading table information for completion of table and column names
You can turn off this feature to get a quicker startup with -A


Database changed
mysql> create table innodb_monitor(a INT) engine=innodb;
Query OK, 0 rows affected (1.04 sec)


mysql> create table innodb_tablespace_monitor(a INT) engine=innodb;
Query OK, 0 rows affected (0.70 sec)


mysql> create table innodb_lock_monitor(a INT) engine=innodb;
Query OK, 0 rows affected (0.36 sec)


mysql> create table innodb_table_monitor(a INT) engine=innodb;
Query OK, 0 rows affected (0.08 sec)


可以通过show engine innodb status命令查看死锁信息
mysql> show engine innodb status \G
*************************** 1. row ***************************
  Type: InnoDB
  Name: 
Status: 
=====================================
2018-05-10 09:17:10 0x7f1fbc21a700 INNODB MONITOR OUTPUT
=====================================
Per second averages calculated from the last 46 seconds
-----------------
BACKGROUND THREAD
-----------------
srv_master_thread loops: 53 srv_active, 0 srv_shutdown, 240099 srv_idle
srv_master_thread log flush and writes: 0
----------
SEMAPHORES
----------
OS WAIT ARRAY INFO: reservation count 2007
OS WAIT ARRAY INFO: signal count 1987
RW-shared spins 3878, rounds 5594, OS waits 1735
RW-excl spins 3, rounds 91, OS waits 4
RW-sx spins 1, rounds 30, OS waits 1
Spin rounds per wait: 1.44 RW-shared, 30.33 RW-excl, 30.00 RW-sx
------------
TRANSACTIONS
------------
Trx id counter 78405
Purge done for trx's n:o < 78404 undo n:o < 10 state: running but idle
History list length 21
LIST OF TRANSACTIONS FOR EACH SESSION:
---TRANSACTION 421249967052640, not started
0 lock struct(s), heap size 1136, 0 row lock(s)
--------
FILE I/O
--------
I/O thread 0 state: waiting for completed aio requests (insert buffer thread)
I/O thread 1 state: waiting for completed aio requests (log thread)
I/O thread 2 state: waiting for completed aio requests (read thread)
.............................................................................
.............................................................................
.............................................................................

来自 “ ITPUB博客 ” ,链接:http://blog.itpub.net/15498/viewspace-2154256/,如需转载,请注明出处,否则将追究法律责任。

请登录后发表评论 登录
全部评论
交流MySQL,MongoDB和Redis技术。 微信或QQ:410294

注册时间:2015-12-07

  • 博文量
    657
  • 访问量
    1382168