服务器之家:专注于服务器技术及软件下载分享
分类导航

Mysql|Sql Server|Oracle|Redis|MongoDB|PostgreSQL|Sqlite|DB2|mariadb|Access|数据库技术|

服务器之家 - 数据库 - Mysql - 深入了解mysql长事务

深入了解mysql长事务

2020-08-10 22:23mysql技术 Mysql

这篇文章主要介绍了mysql长事务的相关资料,帮助大家更好的理解和学习mysql,感兴趣的朋友可以了解下

前言:

本篇文章主要介绍MySQL长事务相关内容,比如说我们开启的一个事务,一直没提交或回滚会怎样呢,出现事务等待情况应该如何处理,本篇文章将给你答案。

注意:本篇文章并不聚焦于谈论事务隔离级别以及相关特性。而是介绍长事务相关危害以及监控处理方法。本文是基于MySQL5.7.23版本,不可重复读(RR)隔离级别所做实验。(语句为\G可以使查询结构显示更易读,但只可以在mysql命令行使用。)

1.什么是长事务

首先我们先要知道什么是长事务,顾名思义就是运行时间比较长,长时间未提交的事务,也可以称之为大事务。这类事务往往会造成大量的阻塞和锁超时,容易造成主从延迟,要尽量避免使用长事务。

下面我将演示下如何开启事务及模拟长事务:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
#假设我们有一张stu_tb表,结构及数据如下
mysql> show create table stu_tb\G
*************************** 1. row ***************************
    Table: stu_tb
Create Table: CREATE TABLE `stu_tb` (
 `increment_id` int(11) NOT NULL AUTO_INCREMENT COMMENT '自增主键',
 `stu_id` int(11) NOT NULL COMMENT '学号',
 `stu_name` varchar(20) DEFAULT NULL COMMENT '学生姓名',
 `create_time` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP COMMENT '创建时间',
 `update_time` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP COMMENT '修改时间',
 PRIMARY KEY (`increment_id`),
 UNIQUE KEY `uk_stu_id` (`stu_id`) USING BTREE
) ENGINE=InnoDB AUTO_INCREMENT=9 DEFAULT CHARSET=utf8 COMMENT='测试学生表'
1 row in set (0.01 sec)
 
mysql> select * from stu_tb;
+--------------+--------+----------+---------------------+---------------------+
| increment_id | stu_id | stu_name | create_time     | update_time     |
+--------------+--------+----------+---------------------+---------------------+
|      1 |  1001 | from1  | 2019-09-15 14:27:34 | 2019-09-15 14:27:34 |
|      2 |  1002 | dfsfd  | 2019-09-15 14:27:34 | 2019-09-15 14:27:34 |
|      3 |  1003 | fdgfg  | 2019-09-15 14:27:34 | 2019-09-15 14:27:34 |
|      4 |  1004 | sdfsdf  | 2019-09-15 14:27:34 | 2019-09-15 14:27:34 |
|      5 |  1005 | dsfsdg  | 2019-09-15 14:27:34 | 2019-09-15 14:27:34 |
|      6 |  1006 | fgd   | 2019-09-15 14:27:34 | 2019-09-15 14:27:34 |
|      7 |  1007 | fgds   | 2019-09-15 14:27:34 | 2019-09-15 14:27:34 |
|      8 |  1008 | dgfsa  | 2019-09-15 14:27:34 | 2019-09-15 14:27:34 |
+--------------+--------+----------+---------------------+---------------------+
8 rows in set (0.00 sec)
 
#显式开启事务,可用begin或start transaction
mysql> start transaction;
Query OK, 0 rows affected (0.00 sec)
 
mysql> select * from stu_tb where stu_id = 1006 for update;
+--------------+--------+----------+---------------------+---------------------+
| increment_id | stu_id | stu_name | create_time     | update_time     |
+--------------+--------+----------+---------------------+---------------------+
|      6 |  1006 | fgd   | 2019-09-15 14:27:34 | 2019-09-15 14:27:34 |
+--------------+--------+----------+---------------------+---------------------+
1 row in set (0.01 sec)
 
 #如果我们不及时提交上个事务,那么这个事务就变成了长事务,当其他会话要操作这条数据时,就会一直等待。

2.如何找到长事务

遇到事务等待问题时,我们首先要做的是找到正在执行的事务。information_schema.INNODB_TRX 表中包含了当前innodb内部正在运行的事务信息,这个表中给出了事务的开始时间,我们可以稍加运算即可得到事务的运行时间。

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
mysql> select t.*,to_seconds(now())-to_seconds(t.trx_started) idle_time from INFORMATION_SCHEMA.INNODB_TRX t \G
*************************** 1. row ***************************
          trx_id: 6168
         trx_state: RUNNING
        trx_started: 2019-09-16 11:08:27
   trx_requested_lock_id: NULL
     trx_wait_started: NULL
        trx_weight: 3
    trx_mysql_thread_id: 11
         trx_query: NULL
    trx_operation_state: NULL
     trx_tables_in_use: 0
     trx_tables_locked: 1
     trx_lock_structs: 3
   trx_lock_memory_bytes: 1136
      trx_rows_locked: 2
     trx_rows_modified: 0
  trx_concurrency_tickets: 0
    trx_isolation_level: REPEATABLE READ
     trx_unique_checks: 1
  trx_foreign_key_checks: 1
trx_last_foreign_key_error: NULL
 trx_adaptive_hash_latched: 0
 trx_adaptive_hash_timeout: 0
     trx_is_read_only: 0
trx_autocommit_non_locking: 0
         idle_time: 170

在结果中idle_time是计算产生的,也是事务的持续时间。但事务的trx_query是NUL,这并不是说事务什么也没执行,一个事务可能包含多个SQL,如果SQL执行完毕就不再显示了。当前事务正在执行,innodb也不知道这个事务后续还有没有sql,啥时候会commit。因此trx_query不能提供有意义的信息。

如果我们想看到这个事务执行过的SQL,看是否可以杀掉长事务,怎么办呢?我们可以联合其他系统表查询得到,具体查询SQL如下:

?
1
2
3
4
5
6
7
8
9
10
mysql> select now(),(UNIX_TIMESTAMP(now()) - UNIX_TIMESTAMP(a.trx_started)) diff_sec,b.id,b.user,b.host,b.db,d.SQL_TEXT from information_schema.innodb_trx a inner join
  -> information_schema.PROCESSLIST b
  -> on a.TRX_MYSQL_THREAD_ID=b.id and b.command = 'Sleep'
  -> inner join performance_schema.threads c ON b.id = c.PROCESSLIST_ID
  -> inner join performance_schema.events_statements_current d ON d.THREAD_ID = c.THREAD_ID;
+---------------------+----------+----+------+-----------+--------+-----------------------------------------------------+
| now()        | diff_sec | id | user | host   | db   | SQL_TEXT                      |
+---------------------+----------+----+------+-----------+--------+-----------------------------------------------------+
| 2019-09-16 14:06:26 |    54 | 17 | root | localhost | testdb | select * from stu_tb where stu_id = 1006 for update |
+---------------------+----------+----+------+-----------+--------+-----------------------------------------------------+

上述结果中diff_sec和上面idle_time表示意思相同,都是代表此事务持续的秒数。SQL_TEXT表示该事务刚执行的SQL。但是呢,上述语句只能查到事务最后执行的SQL,我们知道,一个事务里可能包含多个SQL,那我们想查询这个未提交的事务执行过哪些SQL,是否可以满足呢,答案是结合events_statements_history系统表也可以满足需求。下面语句将会查询出该事务执行过的所有SQL:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
mysql> SELECT
  ->  ps.id 'PROCESS ID',
  ->  ps.USER,
  ->  ps.HOST,
  ->  esh.EVENT_ID,
  ->  trx.trx_started,
  ->  esh.event_name 'EVENT NAME',
  ->  esh.sql_text 'SQL',
  ->  ps.time
  -> FROM
  ->  PERFORMANCE_SCHEMA.events_statements_history esh
  ->  JOIN PERFORMANCE_SCHEMA.threads th ON esh.thread_id = th.thread_id
  ->  JOIN information_schema.PROCESSLIST ps ON ps.id = th.processlist_id
  ->  LEFT JOIN information_schema.innodb_trx trx ON trx.trx_mysql_thread_id = ps.id
  -> WHERE
  ->  trx.trx_id IS NOT NULL
  ->  AND ps.USER != 'SYSTEM_USER'
  -> ORDER BY
  ->  esh.EVENT_ID;
+------------+------+-----------+----------+---------------------+------------------------------+-----------------------------------------------------+------+
| PROCESS ID | USER | HOST   | EVENT_ID | trx_started     | EVENT NAME          | SQL                         | time |
+------------+------+-----------+----------+---------------------+------------------------------+-----------------------------------------------------+------+
|     20 | root | localhost |    1 | 2019-09-16 14:18:44 | statement/sql/select     | select @@version_comment limit 1          |  60 |
|     20 | root | localhost |    2 | 2019-09-16 14:18:44 | statement/sql/begin     | start transaction                  |  60 |
|     20 | root | localhost |    3 | 2019-09-16 14:18:44 | statement/sql/select     | SELECT DATABASE()                  |  60 |
|     20 | root | localhost |    4 | 2019-09-16 14:18:44 | statement/com/Init DB    | NULL                        |  60 |
|     20 | root | localhost |    5 | 2019-09-16 14:18:44 | statement/sql/show_databases | show databases                   |  60 |
|     20 | root | localhost |    6 | 2019-09-16 14:18:44 | statement/sql/show_tables  | show tables                     |  60 |
|     20 | root | localhost |    7 | 2019-09-16 14:18:44 | statement/com/Field List   | NULL                        |  60 |
|     20 | root | localhost |    8 | 2019-09-16 14:18:44 | statement/com/Field List   | NULL                        |  60 |
|     20 | root | localhost |    9 | 2019-09-16 14:18:44 | statement/sql/select     | select * from stu_tb                |  60 |
|     20 | root | localhost |    10 | 2019-09-16 14:18:44 | statement/sql/select     | select * from stu_tb where stu_id = 1006 for update |  60 |
+------------+------+-----------+----------+---------------------+------------------------------+-----------------------------------------------------+------+

从上述结果中我们可以看到该事务从一开始到现在执行过的所有SQL,当我们把该事务相关信息都查询清楚后,我们就可以判定该事务是否可以杀掉,以免影响其他事务造成等待现象。

在这里稍微拓展下,长事务极易造成阻塞或者死锁现象,通常情况下我们可以首先查询 sys.innodb_lock_waits 视图确定有没有事务阻塞现象:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
#假设一个事务执行 select * from stu_tb where stu_id = 1006 for update
#另外一个事务执行 update stu_tb set stu_name = 'wang' where stu_id = 1006
 
mysql> select * from sys.innodb_lock_waits\G
*************************** 1. row ***************************
        wait_started: 2019-09-16 14:34:32
          wait_age: 00:00:03
        wait_age_secs: 3
        locked_table: `testdb`.`stu_tb`
        locked_index: uk_stu_id
         locked_type: RECORD
       waiting_trx_id: 6178
     waiting_trx_started: 2019-09-16 14:34:32
       waiting_trx_age: 00:00:03
   waiting_trx_rows_locked: 1
  waiting_trx_rows_modified: 0
         waiting_pid: 19
        waiting_query: update stu_tb set stu_name = 'wang' where stu_id = 1006
       waiting_lock_id: 6178:47:4:7
      waiting_lock_mode: X
       blocking_trx_id: 6177
        blocking_pid: 20
       blocking_query: NULL
      blocking_lock_id: 6177:47:4:7
     blocking_lock_mode: X
    blocking_trx_started: 2019-09-16 14:18:44
      blocking_trx_age: 00:15:51
  blocking_trx_rows_locked: 2
 blocking_trx_rows_modified: 0
   sql_kill_blocking_query: KILL QUERY 20
sql_kill_blocking_connection: KILL 20

上述结果显示出被阻塞的SQL以及锁的类型,更强大的是杀掉会话的语句也给出来了。但是并没有找到阻塞会话执行的SQL,如果我们想找出更详细的信息,可以使用下面语句:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
mysql> SELECT
  ->  tmp.*,
  ->  c.SQL_Text blocking_sql_text,
  ->  p.HOST blocking_host
  -> FROM
  ->  (
  ->  SELECT
  ->   r.trx_state wating_trx_state,
  ->   r.trx_id waiting_trx_id,
  ->   r.trx_mysql_thread_Id waiting_thread,
  ->   r.trx_query waiting_query,
  ->   b.trx_state blocking_trx_state,
  ->   b.trx_id blocking_trx_id,
  ->   b.trx_mysql_thread_id blocking_thread,
  ->   b.trx_query blocking_query
  ->  FROM
  ->   information_schema.innodb_lock_waits w
  ->   INNER JOIN information_schema.innodb_trx b ON b.trx_id = w.blocking_trx_id
  ->   INNER JOIN information_schema.innodb_trx r ON r.trx_id = w.requesting_trx_id
  ->  ) tmp,
  ->  information_schema.PROCESSLIST p,
  ->  PERFORMANCE_SCHEMA.events_statements_current c,
  ->  PERFORMANCE_SCHEMA.threads t
  -> WHERE
  ->  tmp.blocking_thread = p.id
  ->  AND t.thread_id = c.THREAD_ID
  ->  AND t.PROCESSLIST_ID = p.id \G
*************************** 1. row ***************************
 wating_trx_state: LOCK WAIT
  waiting_trx_id: 6180
  waiting_thread: 19
   waiting_query: update stu_tb set stu_name = 'wang' where stu_id = 1006
blocking_trx_state: RUNNING
  blocking_trx_id: 6177
  blocking_thread: 20
  blocking_query: NULL
 blocking_sql_text: select * from stu_tb where stu_id = 1006 for update
   blocking_host: localhost

上面结果显得更加清晰,我们可以清楚的看到阻塞端及被阻塞端事务执行的语句,有助于我们排查并确认是否可以杀掉阻塞的会话。

3.监控长事务

现实工作中我们需要监控下长事务,定义一个阈值,比如说30s 执行时间超过30s的事务即为长事务,要求记录并告警出来,提醒管理人员去处理。下面给出监控脚本,各位可以参考下,根据需求改动使用:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
#!/bin/bash
# -------------------------------------------------------------------------------
# FileName:  long_trx.sh
# Describe:  monitor long transaction
# Revision:  1.0
# Date:    2019/09/16
# Author:   wang
 
/usr/local/mysql/bin/mysql -N -uroot -pxxxxxx -e "select now(),(UNIX_TIMESTAMP(now()) - UNIX_TIMESTAMP(a.trx_started)) diff_sec,b.id,b.user,b.host,b.db,d.SQL_TEXT from information_schema.innodb_trx a inner join
information_schema.PROCESSLIST b
on a.TRX_MYSQL_THREAD_ID=b.id and b.command = 'Sleep'
inner join performance_schema.threads c ON b.id = c.PROCESSLIST_ID
inner join performance_schema.events_statements_current d ON d.THREAD_ID = c.THREAD_ID;" | while read A B C D E F G H
do
 if [ "$C" -gt 30 ]
   then
   echo $(date +"%Y-%m-%d %H:%M:%S")
   echo "processid[$D] $E@$F in db[$G] hold transaction time $C SQL:$H"
 fi
done >> /tmp/longtransaction.txt

简单说明一下,这里的-gt 30是30秒钟的意思,只要超过了30秒钟就认定是长事务,可以根据实际需要自定义。将该脚本加入定时任务中即可执行。

总结:

本文主要介绍了长事务相关内容,怎样找到长事务,怎么处理长事务,如何监控长事务。可能有些小伙伴对事务理解还不多,希望这篇文章对你有所帮助。由于本篇文章列出的查询事务相关语句较多,现总结如下:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
# 查询所有正在运行的事务及运行时间
select t.*,to_seconds(now())-to_seconds(t.trx_started) idle_time from INFORMATION_SCHEMA.INNODB_TRX t \G
 
# 查询事务详细信息及执行的SQL
select now(),(UNIX_TIMESTAMP(now()) - UNIX_TIMESTAMP(a.trx_started)) diff_sec,b.id,b.user,b.host,b.db,d.SQL_TEXT from information_schema.innodb_trx a inner join information_schema.PROCESSLIST b
on a.TRX_MYSQL_THREAD_ID=b.id and b.command = 'Sleep'
inner join performance_schema.threads c ON b.id = c.PROCESSLIST_ID
inner join performance_schema.events_statements_current d ON d.THREAD_ID = c.THREAD_ID;
 
# 查询事务执行过的所有历史SQL记录
SELECT
 ps.id 'PROCESS ID',
 ps.USER,
 ps.HOST,
 esh.EVENT_ID,
 trx.trx_started,
 esh.event_name 'EVENT NAME',
 esh.sql_text 'SQL',
 ps.time
FROM
 PERFORMANCE_SCHEMA.events_statements_history esh
 JOIN PERFORMANCE_SCHEMA.threads th ON esh.thread_id = th.thread_id
 JOIN information_schema.PROCESSLIST ps ON ps.id = th.processlist_id
 LEFT JOIN information_schema.innodb_trx trx ON trx.trx_mysql_thread_id = ps.id
WHERE
 trx.trx_id IS NOT NULL
 AND ps.USER != 'SYSTEM_USER'
ORDER BY
 esh.EVENT_ID;
 
 # 简单查询事务锁
 select * from sys.innodb_lock_waits\G
 
 # 查询事务锁详细信息
 SELECT
 tmp.*,
 c.SQL_Text blocking_sql_text,
 p.HOST blocking_host
FROM
 (
 SELECT
  r.trx_state wating_trx_state,
  r.trx_id waiting_trx_id,
  r.trx_mysql_thread_Id waiting_thread,
  r.trx_query waiting_query,
  b.trx_state blocking_trx_state,
  b.trx_id blocking_trx_id,
  b.trx_mysql_thread_id blocking_thread,
  b.trx_query blocking_query
 FROM
  information_schema.innodb_lock_waits w
  INNER JOIN information_schema.innodb_trx b ON b.trx_id = w.blocking_trx_id
  INNER JOIN information_schema.innodb_trx r ON r.trx_id = w.requesting_trx_id
 ) tmp,
 information_schema.PROCESSLIST p,
 PERFORMANCE_SCHEMA.events_statements_current c,
 PERFORMANCE_SCHEMA.threads t
WHERE
 tmp.blocking_thread = p.id
 AND t.thread_id = c.THREAD_ID
 AND t.PROCESSLIST_ID = p.id \G

以上就是深入了解mysql长事务的详细内容,更多关于mysql长事务的资料请关注服务器之家其它相关文章!

原文链接:https://cloud.tencent.com/developer/article/1511089

延伸 · 阅读

精彩推荐