见官方文档:http://dev.mysql.com/doc/refman/5.1/zh/storage-engines.html#innodb-lock-modes
可直接在mysql命令行执行:show engine innodb status\G;
查看造成死锁的sql语句,分析索引情况,然后优化sql
然后
show processlist;
kill processid;
另外可以打开慢查询日志,linux下打开需在my.cnf的[mysqld]里面加上以下内容:
long_query_time = 2
log-slow-queries = /usr/local/mysql/mysql-slow.log
最近在项目开发过程中,碰到了数据库的死锁问题,在解决问题的过程中,加深了对MySQLInnoDB引擎锁机制的理解。
我们使用Show innodb status检查引擎状态时,发现了死锁问题:
*** (1) TRANSACTION:
TRANSACTION 0 677833455, ACTIVE 0 sec, process no 11393, OS threadid 278546 starting index read
mysql tables in use 1, locked 1
LOCK WAIT 3 lock struct(s), heap size 320
MySQL thread id 83, query id 162348740 dcnet03 dcnet Searching rowsfor update
update TSK_TASK set STATUS_ID=1064,UPDATE_TIME=now () whereSTATUS_ID=1061 and MON_TIME<date_sub(now(), INTERVAL30 minute)
*** (1) WAITING FOR THIS LOCK TO BE GRANTED:
RECORD LOCKS space id 0 page no 849384 n bits 208 index `PRIMARY`of table `dcnet_db/TSK_TASK` trx id 0 677833455 lock_mode X locksrec but not gap waiting
Record lock, heap no 92 PHYSICAL RECORD: n_fields 11; compactformat; info bits 0
0: len 8; hex 800000000097629c; asc
b ;; 1: len 6; hex00002866eaee; asc (f ;; 2: len7; hex 00000d40040110; asc @ ;; 3: len 8; hex 80000000000050b2;asc P ;; 4:len 8; hex 800000000000502a; asc P*;; 5: len 8; hex8000000000005426; asc T&;; 6: len 8; hex800012412c66d29c; asc A,f ;; 7: len 23; hex75706c6f6164666972652e63 6f6d2f6 8616e642e706870; asc xxx.com/;; 8:len 8; hex 800000000000042b; asc +;; 9: len4; hex 474bfa2b; asc GK +;; 10: len 8; hex 8000000000004e24;asc N$;;
*** (2) TRANSACTION:
TRANSACTION 0 677833454, ACTIVE 0 sec, process no 11397, OS threadid 344086 updating or deleting, thread declared inside InnoDB499
mysql tables in use 1, locked 1
3 lock struct(s), heap size 320, undo log entries 1
MySQL thread id 84, query id 162348739 dcnet03 dcnetUpdating
update TSK_TASK set STATUS_ID=1067,UPDATE_TIME=now () where ID in(9921180)
*** (2) HOLDS THE LOCK(S):
RECORD LOCKS space id 0 page no 849384 n bits 208 index `PRIMARY`of table `dcnet_db/TSK_TASK` trx id 0 677833454 lock_mode X locksrec but not gap
Record lock, heap no 92 PHYSICAL RECORD: n_fields 11; compactformat; info bits 0
0: len 8; hex 800000000097629c; asc b ;; 1: len 6; hex00002866eaee; asc (f ;; 2: len7; hex 00000d40040110; asc @ ;; 3: len 8; hex 80000000000050b2;asc P ;; 4:len 8; hex 800000000000502a; asc P*;; 5: len 8; hex8000000000005426; asc T&;; 6: len 8; hex800012412c66d29c; asc A,f ;; 7: len 23; hex75706c6f6164666972652e63 6f6d2f6 8616e642e706870; ascuploadfire.com/hand.php;; 8: len 8; hex 800000000000042b;asc +;; 9: len 4; hex 474bfa2b; asc GK +;; 10: len8; hex 8000000000004e24; asc N$;;
*** (2) WAITING FOR THIS LOCK TO BE GRANTED:
RECORD LOCKS space id 0 page no 843102 n bits 600 index`KEY_TSKTASK_MONTIME2` of table `dcnet_db/TSK_TASK` trx id 0677833454 lock_mode X locks rec but not gap waiting
Record lock, heap no 395 PHYSICAL RECORD: n_fields 3; compactformat; info bits 0
0: len 8; hex 8000000000000425; asc %;; 1: len8; hex 800012412c66d29c; asc A,f ;; 2: len 8; hex 800000000097629c;asc b;;
*** WE ROLL BACK TRANSACTION (1)
该死锁问题涉及TSK_TASK表,该表用于保存系统监测任务,相关字段及索引如 下:
ID:主键;
MON_TIME:监测时间;
STATUS_ID:任务状态;
索 引:KEY_TSKTASK_MONTIME2 (STATUS_ID, MON_TIME)。
经分析,涉及的两条语句应该不会 涉及相同的TSK_TASK记录,那为什么会造成死锁呢?
查询MySQL官网文档,发现这跟MySQL的索引机制有关。MySQL的InnoDB引擎是行级锁,我原来的理解是直接对记录进行锁定,实际上不是这样的,其要点如下:
不是对记录进行锁定,而是对索引 进行锁定;
在UPDATE、DELETE操作时,MySQL不仅锁定WHERE条件扫描过的所有索引记录,而且会锁定相邻的键值,即所谓的next-key locking;
如语句UPDATE TSK_TASK SET UPDATE_TIME = NOW() WHERE ID>10000会锁定所有主键大于等于1000的所有记录,在该语句完成之前,你就不能对主键等于10000的记录进行操作;
当 非簇索引(non-cluster index)记录被锁定时,相关的簇索引(clusterindex)记录也需要被锁定才能完成相应的操作。
再 分析一下发生问题的两条SQL语句,就不难找到问题所在了:
当“update TSK_TASK set STATUS_ID=1064,UPDATE_TIME=now () whereSTATUS_ID=1061 and MON_TIME<date_sub(now(), INTERVAL30 minute)”执行时,MySQL会使用KEY_TSKTASK_MONTIME2索引,因此首先锁定相关的索引记录,因为KEY_TSKTASK_MONTIME2是非簇索引,为执行该语句,MySQL还会锁定簇索引(主键索引)。
假设“update TSK_TASK set STATUS_ID=1067,UPDATE_TIME=now () where IDin (9921180)”几乎同时执行时,本语句首先锁定簇索引(主键),由于需要更新STATUS_ID的值,所以还需要锁定KEY_TSKTASK_MONTIME2的某些索引记录。
这样第一条语句锁定了KEY_TSKTASK_MONTIME2的记录,等待主键索引,而第二条语句则锁定了主键索引记录,而等待KEY_TSKTASK_MONTIME2的记录,这样死锁就产生了。
我们通过拆分 第一条语句解决了死锁问题:即先查出符合条件的ID:select ID from TSK_TASK whereSTATUS_ID=1061 and MON_TIME < date_sub(now(),INTERVAL 30 minute);然后再更新状态:update TSK_TASK set STATUS_ID=1064where ID in (….)
这样就不会产生索引的竞争问题,死锁问题就 解决了