MySQL essay

/* View all permissions of the user */
show grants for 'root'@'localhost';
/* Set to match all permissions */
 grant all privileges on *.* to 'root'@'%';
/* Refresh the permissions */
 flush privileges;

install mysql

bin/mysqld --defaults-file=/usr/local/mysql2/my.cnf --initialize --user=mysql
# Log in to the server after getting the password   Set global access and change password , At this point the installation is complete 

mysql Loading order of configuration files

file name	Purpose
/etc/my.cnf	Global options
/etc/mysql/my.cnf	Global options
SYSCONFDIR/my.cnf	Global options
$MYSQL_HOME/my.cnf	Server specific options （ Server only ）
defaults-extra-file	use Specified file --defaults-extra-file, If any
~/.my.cnf	User specific options
~/.mylogin.cnf	User specific login Path Options （ Client only ）

mysql Database summary

mysql A database is a system database , Contains storage MySQL A table of information required for the server to run .
Table functions are divided into the following categories :

• Grant system table Grant System Tables
• Object information system table Object Information System Tables
• Record system table Log System Tables
• Server side help system table Server-Side Help System Tables
• Time zone system table Time Zone System Tables
• Copy system tables Replication System Tables
• Optimizer system tables Optimizer System Tables
• Miscellaneous system table Miscellaneous System Tables

The point is Copy the tables of the system , In order to facilitate our future study Database Read / write separation .
gtid_executed： Used to store GTID A table of values , Use InnoDB Storage engine .
ndb_binlog_index: NDB Cluster Copied binary log information .
slave_master_info, slave_relay_log_info, slave_worker_info: Used to store replication information on the replica server .

mysql Server log

Log Type	Information Written to Log
Error log	start-up 、 To run or stop mysqld Log of problems encountered during
General query log	Established client connections and statements received from clients
Binary log	A statement that changes data
Relay log	Data received from the replication source server
Slow query log	Execution time exceeded long_query_time A few seconds of inquiry
DDL log (metadata log)	DDL Metadata operations executed by statement

Turn on the general log function

log_output=FILE
general_log=ON
general_log_file=general.log

Focus on binary log journal

binary log Format of log

• be based on SQL sentence binlog-format=STATEMENT
• Line based logging binlog-format=ROW
• Mixed logging binlog-format=MIXED

change bin log Format considerations

• When master-slave synchronization is enabled , If Master slave asynchronous modification , You're going to report a mistake
• Use Read submitted and Read uncommitted transaction isolation level , Only use Line based log format

Set different log formats for different storage engines

Storage engine	ROW	STATEMENT
ARCHIVE	Yes	Yes
BLACKHOLE	Yes	Yes
CSV	Yes	Yes
EXAMPLE	Yes	No
FEDERATED	Yes	Yes
HEAP	Yes	Yes
InnoDB	Yes	yes , When transaction isolation level is REPEATABLE READ or when SERIALIZABLE; Otherwise no .
MyISAM	Yes	Yes
MERGE	Yes	Yes
NDB	Yes	No

bin log Opportunity to generate new bin log journal , Small scale from 1 Start to grow

• The server starts or restarts
• Server refresh log
• The current log file size has reached max_binlog_size

bin log The logging action is in uncommitted transactions , All changes will be cached , until The server received a commit statement .mysqld Write the entire transaction to the binary log before execution . When the thread processing the transaction starts , It will binlog_cache_size Allocate a buffer for buffer statements . If the statement is greater than this value , Then the thread will open a temporary file to store transactions . Delete temporary files at the end of the thread .

binlog To configure

server-id=1
log-bin=/usr/local/mysql/mysql-files/binLog
log_bin_index=/usr/local/mysql/mysql-files/binLogIndex
binlog_cache_size=4096
max_binlog_cache_size=4096
sync_binlog=1
binlog_format=MIXED

/*  Set the global variable on bin log*/
SET GLOBAL log_bin = ON;

Slow query log

# Slow query log 
slow_query_log=ON
long_query_time=10
min_examined_row_limit=10000
slow_query_log_file=slow.log
log_output=FILE

Backup

# Back up all databases 
./bin/mysqldump --all-databases -uroot -p > dump.sql
# Back up the specified database 
 mysqldump --databases db1 db2 db3 > dump.sql
 # Refresh and back up data 
 mysqldump --single-transaction --flush-logs --master-data=2 --all-databases > backup_sunday_1_PM.sql
 # Back up and delete   Binary 
mysqldump --single-transaction --flush-logs --master-data=2 --all-databases --delete-master-logs > backup_sunday_1_PM.sql

 #  Import sql
 mysql < dump.sql
#sql  If there is no statement to create a table in 
mysqladmin create db1
mysql db1 < dump.sql
# or 
CREATE DATABASE IF NOT EXISTS db1;
USE db1;
source dump.sql

/* Check all the logs */
SHOW BINARY LOGS;
/* View the current log  */
SHOW MASTER STATUS;

flush logs;

Full recovery

/* Incremental backup recovery   Mode one */
mysqlbinlog binlog.000001 binlog.000002 | mysql -u root -p
/* Incremental backup recovery   Mode two */
mysqlbinlog binlog.000001 >  /tmp/statements.sql
mysqlbinlog binlog.000002 >> /tmp/statements.sql
mysql -u root -p -e "source /tmp/statements.sql"

Point in time recovery using event location

mysqlbinlog   --start-datetime="2020-05-27 12:59:00" --stop-datetime="2020-05-27 13:06:00" \
  --verbose /var/lib/mysql/bin.123456 | grep -C 12 "DROP TABLE"

#  Restore to the specified event point 
mysqlbinlog --start-position=4 --stop-position=1548 binLog.000006 | mysql -u root -p

# Start from the specified event point 
mysqlbinlog --start-position=3137 binLog.000006 | mysql -u root -p

Tips Try to do it manually before performing the backup flush logs; Generate a new backup file , avoid binLog A duplicate record of

Clustered index and secondary index

Every InnoDB Tables have a special index , It's called a clustered index , Used to store row data . Usually , A clustered index is synonymous with a primary key .

1. Yes PRIMARY KEY,InnoDB Use it as a clustered index .
2. If you don't PRIMARY KEY,InnoDB Use the first UNIQUE Indexes , All key columns are defined as NOT NULL.
3. If the table has no index PRIMARY KEY Or not suitable Of UNIQUE Indexes , be InnoDB Generate a hidden clustered index , The index is in GEN_CLUST_INDEX Include lines ID Name the composite column of values .

Indexes other than clustered indexes are called secondary indexes . stay InnoDB in , Each record in the secondary index contains the primary key column of the row , And the columns specified for the secondary index . InnoDB Use this primary key value to search the clustered index for rows .

Copy

Copy based on binary log file location

1. The source server Create a copy permission account and permission

CREATE USER 'repl'@'%' IDENTIFIED BY '123456';
GRANT REPLICATION SLAVE ON *.* TO 'repl'@'%';

2. Back up the log files of the source server

./bin/mysqldump --all-databases -uroot -p > dump.sql

3. Install replica server
4. Configure replica server configuration

[mysqld]
port=3307
datadir=/usr/local/mysql2/mysql-files
socket=/tmp/mysql2.sock
risks
symbolic-links=0
#server id
server-id=2
[mysqld_safe]
log-error=/usr/local/mysql2/mysql-files/mariadb.log
pid-file=/usr/local/mysql2/mysql-files/mariadb.pid
!includedir /etc/my.cnf.d

5. Start the slave server with the specified configuration file
6. Configure the source server information on the slave server ,post And documents Can be viewed from the source server .

   /* Can be viewed from the source server */
   SHOW MASTER STATUS;
   

   /* Configure the source server information on the slave server ,post  And documents  */
   CHANGE MASTER TO
        MASTER_HOST='localhost',
    			MASTER_PORT=3306,
        MASTER_USER='repl',
        MASTER_PASSWORD='123456',
        MASTER_LOG_FILE='binLog.000012',
        MASTER_LOG_POS=154;
   

7. Import full data for the source server
8. Start the source server SLAVE Threads

   START SLAVE;
   

Complete master-slave configuration .