MySQL主從復(fù)制實(shí)戰(zhàn)-詳細(xì)說(shuō)明基于日志點(diǎn)的復(fù)制代碼案例

基于日志點(diǎn)的復(fù)制

1、在主庫(kù)與從庫(kù)上建立專(zhuān)用的復(fù)制賬號(hào)

MariaDB [employees]> create user 'repl'@'172.%' identified by '123456';

注意在生產(chǎn)上的密碼必須依照相關(guān)規(guī)范以達(dá)到一定的密碼強(qiáng)度, 并且規(guī)定在從庫(kù)上的特定網(wǎng)段上才能訪問(wèn)主庫(kù)

2、在主庫(kù)與從庫(kù)上授予復(fù)制權(quán)限

MariaDB [employees]> grant replication slave on *.* to 'repl'@'172.%';

3、配置主庫(kù)

注意啟用二進(jìn)制日志需要重啟服務(wù), 而server_id是一個(gè)動(dòng)態(tài)參數(shù), 可以結(jié)合命令行與配置文件以達(dá)到免重啟的持久化配置. 注意server_id在集群中是唯一的.

[mysqld]
log_bin = /var/log/mysql/mariadb-bin
log_bin_index = /var/log/mysql/mariadb-bin.index
binlog_format = row
server_id = 101

NOTE: 把日志與數(shù)據(jù)分開(kāi)是個(gè)好習(xí)慣, 最好能放到不同的數(shù)據(jù)分區(qū)

4、配置從庫(kù)

選項(xiàng)log_slave_update決定是否把中繼日志relay_log存放到本機(jī)的binlog中, 如果是配置鏈路復(fù)制, 那么該選項(xiàng)必填. 注意server_id在集群中是唯一的.

[mysqld]
# replication
log_bin = /var/log/mysql/mariadb-bin
log_bin_index = /var/log/mysql/mariadb-bin.index
server_id = 102
# slaves
relay_log    = /var/log/mysql/relay-bin
relay_log_index  = /var/log/mysql/relay-bin.index
relay_log_info_file  = /var/log/mysql/relay-bin.info
log_slave_updates = ON
read_only

5、初始化從庫(kù)的數(shù)據(jù)

此處使用mysqldump在主庫(kù)上進(jìn)行備份, 在生產(chǎn)上建議大家用xtrabackup進(jìn)行無(wú)鎖的熱備(基于innodb引擎).

備份主庫(kù)上的employees數(shù)據(jù)庫(kù)的數(shù)據(jù)

代碼如下:

mysqldump --single-transaction --master-data=1 --triggers --routines --databases employees -u root -p >> backup.sql

將備份文件backup.sql通過(guò)scp或者docker volume卷掛載到從服務(wù)器上, 并且導(dǎo)入至從庫(kù)中

mysql -u root -p < backup.sql

6、啟動(dòng)復(fù)制鏈路

現(xiàn)有master@172.20.0.2和slave@172.20.0.3, 并且已經(jīng)通過(guò)mysqldump將數(shù)據(jù)同步至從庫(kù)slave中. 現(xiàn)在在從服務(wù)器slave上配置復(fù)制鏈路

MariaDB [(none)]> CHANGE MASTER TO MASTER_HOST='master', 
MASTER_USER='repl', 
MASTER_PASSWORD='123456', 
MASTER_LOG_FILE='mariadb-bin.000029', 
MASTER_LOG_POS=516;
Query OK, 0 rows affected (0.02 sec)

在從庫(kù)上啟動(dòng)復(fù)制鏈路

MariaDB [(none)]> start slave;
Query OK, 0 rows affected (0.01 sec)

7、在從庫(kù)上檢查slave狀態(tài)

Slave_IO_Running與Slave_SQL_Running必須為YES, 如果出現(xiàn)錯(cuò)誤須詳細(xì)閱讀Last_IO_Error或Last_SQL_Error的提示信息

MariaDB [(none)]> show slave status\G
*************************** 1. row ***************************
        Slave_IO_State: Waiting for master to send event
         Master_Host: master
         Master_User: repl
         Master_Port: 3306
        Connect_Retry: 60
       Master_Log_File: mariadb-bin.000029
     Read_Master_Log_Pos: 516
        Relay_Log_File: relay-bin.000002
        Relay_Log_Pos: 539
    Relay_Master_Log_File: mariadb-bin.000029
       Slave_IO_Running: Yes
      Slave_SQL_Running: Yes
       Replicate_Do_DB:
     Replicate_Ignore_DB:
      Replicate_Do_Table:
    Replicate_Ignore_Table:
   Replicate_Wild_Do_Table:
 Replicate_Wild_Ignore_Table:
          Last_Errno: 0
          Last_Error:
         Skip_Counter: 0
     Exec_Master_Log_Pos: 516
       Relay_Log_Space: 831
       Until_Condition: None
        Until_Log_File:
        Until_Log_Pos: 0
      Master_SSL_Allowed: No
      Master_SSL_CA_File:
      Master_SSL_CA_Path:
       Master_SSL_Cert:
      Master_SSL_Cipher:
        Master_SSL_Key:
    Seconds_Behind_Master: 0
Master_SSL_Verify_Server_Cert: No
        Last_IO_Errno: 0
        Last_IO_Error:
        Last_SQL_Errno: 0
        Last_SQL_Error:
 Replicate_Ignore_Server_Ids:
       Master_Server_Id: 101
        Master_SSL_Crl:
      Master_SSL_Crlpath:
          Using_Gtid: No
         Gtid_IO_Pos:
   Replicate_Do_Domain_Ids:
 Replicate_Ignore_Domain_Ids:
        Parallel_Mode: conservative
1 row in set (0.00 sec)

8、在主庫(kù)檢查dump線程

檢測(cè)是否已經(jīng)正確啟動(dòng)binlog dump線程

MariaDB [(none)]> show processlist \G
*************************** 1. row ***************************
   Id: 7
  User: root
  Host: 172.20.0.1:41868
   db: employees
 Command: Sleep
  Time: 56
  State:
  Info: NULL
Progress: 0.000
*************************** 2. row ***************************
   Id: 10
  User: repl
  Host: 172.20.0.3:45974
   db: NULL
 Command: Binlog Dump
  Time: 246
  State: Master has sent all binlog to slave; waiting for binlog to be updated
  Info: NULL
Progress: 0.000

可以看到row 2上有Command為Binlog Dump的命令被啟動(dòng), 證明復(fù)制線程已經(jīng)被成功啟動(dòng)

9、總結(jié)

優(yōu)點(diǎn)

1. 技術(shù)成熟, BUG相對(duì)較少

2. 對(duì)SQL查詢(xún)沒(méi)有任何限制, 如基于GTID復(fù)制時(shí)不是所有SQL都可以使用
   

缺點(diǎn)

1. 故障轉(zhuǎn)移時(shí)重新獲取新主的日志偏移量較為困難

在一主多從環(huán)境下, 若舊master宕機(jī)后在集群中選舉出新master, 其他的從庫(kù)要對(duì)這個(gè)新的master進(jìn)行重新同步, 由于每個(gè)DB的binlog都是獨(dú)立存在, 所以很難找出開(kāi)始同步的日志點(diǎn)

以上就是MySQL主從復(fù)制實(shí)戰(zhàn)-詳解基于日志點(diǎn)的復(fù)制代碼實(shí)例的詳細(xì)內(nèi)容，更多請(qǐng)關(guān)注php中文網(wǎng)其它相關(guān)文章！