mysql_problem handling

Table of contents

5. MySQL's official website description about locks

A mysql FAQ

1. Server returns invalid timezone caused by wrong time zone

Solution: mysql> set global time_zone= '+8:00';

Production configuration generally adds a time zone to the url, such as:
jdbc:mysql://ip:port/xk?useUnicode=true&characterEncoding=utf-8&serverTimezone=Asia/Shanghai

2. Data too long for column table field definition is too short, increase the field length or type

alter table table_name modify column_name varchar(100) comment 'check type';
When the table is large, it is recommended to use the gh-ost method for lock-free table update; or the temporary table method?

3.1064 syntax

If you execute the stored procedure: escape the content
delimiter // stored procedure //
If it is a sql statement: it is generally caused by special characters

4 ORA-00907 error in ibatis Use attribute plus [] to solve (ids is a comma-separated string)

<iterate property="ids" conjunction="," open="(" close=")">

　　#ids[]#

</iterate>

Two, commonly used

String concatenation:

update table_name set column_name=concat(column_name,'xxxx') where id ='xx';

date formatting

date_format(create_time, '%Y-%m-%d %H:%i:%S') create_time

create_time datetime default now() comment 'Creation time',
create_by varchar(60) default 'system' not null comment 'creator',
update_time datetime default now() on update now() comment 'update time',
update_by varchar(60) comment 'update founder'

Current time minus seconds minutes, hours

select * from check_config where create_time<=now() -interval 6 second

select * from check_config where create_time<=now() -interval 6 minute

select * from check_config where create_time<=now() -interval 6 hour

tips:

If you compare numbers in mybatis mybatis-plus xml, it is recommended to use .toString()

<if test="status!=null and status=='0'.toString()"> status=#{status} </if>

Combined class is judged by size
<if test="xxlist!=null and xxlist.size()>0"></if>

Character set view

show variables like '%char%';

Check the number of connections. If you encounter insufficient connections, you can use jstack -l pid > /tmp/xx.txt to download: sz /tmp/xx.txt to see if the specific thread pool is too large

#Currently all connections
show full processlist;
#Query the maximum number of connections
show variables like 'max_connections';
-- Set the maximum number of connections, set as needed
set global max_connections=200;
show variables like '%max_connections%';

Modify table fields or delete

alter table table_name add column xx varchar(200) comment 'check model' after xx0;
alter table table_name rename column xx  to xx2;

alter table table_name drop xx;

Index creation or deletion

create index idx_xx on table_name(xx);
create unique index  uq_xx on table_name(xx);

drop index idx_xx on table_name;

insert statement

Ignore the insert, and ignore it if a primary key conflict is encountered
INSERT IGNORE INTO  table_name ...
Repeated updates, as few as possible inserting a large amount of data will cause table locks
INSERT INTO  check_config on duplicate  key update xx=values(xx)
insert into ... select insert method (too much data leads to table lock, a small amount of data backup can be used)
insert into xx  select * from xx2  on duplicate key update column_name=values(column_name);

Special symbol escape processing: use <![CDATA[content]]> in mybatis and mybati-splus

idea simplifies sql shortcut keys (the premise is that it can be configured through the navigation bar on the right side of idea, if you encounter mysql-connector problems, you can download them first through maven and then solve them by adding them locally) as shown in the figure

Similar to the shortcut key commands that can be used in plsql: such as SF space, it will generate select * from

Organizing tables: (regular delete data cannot clean up the data stored in the data storage, but you can use the structuring table statement, but it is recommended not to operate less, you can search for keywords to query the problems introduced by structuring tables) For example, it may be useful to clean up historical data

optimize table user_address;

delete table

truncate table xx;
drop table if exists xx;

Check the lock

show engine innodb status;
show status like 'innodb_row_lock%';
select * from information_schema.innodb_trx;
SELECT * from information_schema.`PROCESSLIST`;

mybati-plus integrates springboot maven

<dependency>
    <groupId>com.baomidou</groupId>
    <artifactId>mybatis-plus-boot-starter</artifactId>
    <version>3.5.1</version>
</dependency>
<dependency>
    <groupId>mysql</groupId>
    <artifactId>mysql-connector-java</artifactId>
    <version>8.0.28</version>
</dependency>

3. Cautions

Lock table problem: It is not recommended to use the update xx select xx from xx statement on the main business table to update

When a certain in field is updated, the less the primary key content control, the better, and try not to exceed 200-300?: update xx set xx=xxx where id in() Syntax The less data in in, the better, and more will cause lock upgrade problems?

Insert leads to improper insertion and also causes deadlock problems (unique index and joint index). In addition, some large transactions that have not been submitted for repeated consumption and consumption will occasionally cause insert deadlock problems. For insert deadlock problems, you can refer to other articles on the Internet or the official website of mysql.

Check the lock timeout configuration: SHOW VARIABLES LIKE 'innodb_lock_wait_timeout';

Index: Commonly used fields, it is recommended to increase the status value to distinguish data, not as many as possible Make a "trade-off between time and space" Time for space, space for time

In terms of query efficiency: avoid calculations on index columns, combine index leftmost matching rules, and add indexes to other indexes that should be added

Mysql query back to the table: set the primary key through the index, and find the specific data through the primary key; if the query column is an index column, then it is not necessary to set it as data twice through the query back to the table (that is, the part of the explain explanation plan in the index coverage index optimization)

Use the following test code to see the specific situation:

SELECT user_name FROM user_address where user_name='u2';
SELECT * FROM user_address where user_name='u2';

Do not go back to the table query, direct index coverage (Using index)

Index coverage is not used, and the table is queried

Special primary key id: if it is an auto-incrementing primary key, it is both data continuous and physical continuous

If there is a non-auto-increment primary key, there will be page splitting : for details, please refer to other articles

Four, slow sql positioning

Mainly to open the slow sql log, and if you know where it is, use explain to explain and execute it to see the specific index

windows environment

show variables like '%slow_query_log%';
set global slow_query_log = on;
# DESKTOP-R3DF1QE-slow.log
set global slow_query_log_file="C:/data/mysql/slow_query_log.log";
show variables like '%long_query_time%';
#  10.000000 默认10秒
set global long_query_time = 1.000000;

Such as slow sql test of local query:

select sleep(2);

C:\Program Files\MySQL\MySQL Server 8.0\bin\mysqld.exe, Version: 8.0.18 (MySQL Community Server - GPL). started with:
TCP Port: 3306, Named Pipe: MySQL
Time Id Command Argument
# Time: 2022-09-28T05:45:03.937414Z
# User@Host: root[root] @ localhost [127.0.0.1] Id: 8
# Query_time: 11.003896 Lock_time: 0.000000 Rows_sent: 1 Rows_examined: 1
use xk;
SET timestamp=1664343892;
/* ApplicationName=IntelliJ IDEA 2022.1 */ select sleep(11);
# Time: 2022-09-28T08:44:23.059353Z
# User@Host: root[root] @ localhost [127.0.0.1] Id: 16
# Query_time: 2.001915 Lock_time: 0.000000 Rows_sent: 1 Rows_examined: 1
SET timestamp=1664354661;
/* ApplicationName=IntelliJ IDEA 2022.1 */ select sleep(2);

linux environment

After logging out, log in again to view

[root@localhost ~]# mysql -uroot -p
Enter password:

mysql> show variables like '%slow_query_log%';
+---------------------+-----------------------------------+
| Variable_name       | Value                             |
+---------------------+-----------------------------------+
| slow_query_log      | ON                                |
| slow_query_log_file | /var/lib/mysql/localhost-slow.log |
+---------------------+-----------------------------------+

mysql> set global slow_query_log = ON;
Query OK, 0 rows affected (0.00 sec)

mysql> show variables like '%long_query_time%';
+-----------------+----------+
| Variable_name   | Value    |
+-----------------+----------+
| long_query_time | 1.000000 |
+-----------------+----------+
1 row in set (0.00 sec)
mysql> set global long_query_time = 1.000000;
Query OK, 0 rows affected (0.00 sec)


mysql> exit;

再次进入
[root@localhost ~]# mysql -uroot -p
Enter password:

mysql> show variables like '%long_query_time%';
+-----------------+----------+
| Variable_name   | Value    |
+-----------------+----------+
| long_query_time | 1.000000 |
+-----------------+----------+
1 row in set (0.00 sec)


mysql> show variables like '%slow_query_log%';
+---------------------+-----------------------------------+
| Variable_name       | Value                             |
+---------------------+-----------------------------------+
| slow_query_log      | ON                                |
| slow_query_log_file | /var/lib/mysql/localhost-slow.log |
+---------------------+-----------------------------------+
2 rows in set (0.00 sec)

If it is some cloud platform, it will have the function of exporting mysql files

How much is considered slow SQL depends on the actual business

5. MySQL's official website description about locks

MySQL :: MySQL 8.0 Reference Manual :: 15.7.1 InnoDB Locking