内容阅读

在知识库中专注阅读，并随时返回相关工具与课程

备份恢复 & 迁移20 分钟阅读

MySQL 8 clone 插件快速重建主从集群

今天发现有一套 MySQL 8 数据库主从不同步了，报错如下： sql mysql show slave status \G; -- 错误内容 LastErrno: 1032 LastError: Coordinator stopped because there were errors in the workers. The most recent failure being: Worker 1

2026年1月8日

mysql

DBA 实战课程

系统学习 Oracle DBA

2738+ 篇实战文章
1500+ 步骤截图
10 大核心模块

开通会员 ¥128/月起

查看完整课程

返回博客列表

备份恢复 & 迁移20 分钟阅读

MySQL 8 clone 插件快速重建主从集群

2026年1月8日

mysql

前言

本章目标：掌握本章核心知识点 前置要求：完成前序章节学习 预计时长：60 分钟

今天发现有一套 MySQL 8 数据库主从不同步了，报错如下：

SQL

1mysql> show slave status \G;
2-- 错误内容
3Last_Errno: 1032
4Last_Error: Coordinator stopped because there were error(s) in the worker(s). The most recent failure being: Worker 1 failed executing transaction 'ANONYMOUS' at master log mysql-bin.003633, end_log_pos 603058578. See error log and/or performance_schema.replication_applier_status_by_worker table for more details about this failure or others, if any.
5
6-- 根据提示消息查询详细报错
7mysql> SELECT * FROM performance_schema.replication_applier_status_by_worker WHERE LAST_ERROR_NUMBER = 1032 \G;
8-- 错误内容
9LAST_ERROR_TIMESTAMP: 2025-09-06 20:08:06.998256
10...
11LAST_ERROR_MESSAGE: Worker 1 failed executing transaction 'ANONYMOUS' at master log mysql-bin.003633, end_log_pos 603058578; 
12Could not execute Update_rows event on table ci_sch***.qrtz_***_triggers; 
13Can't find record in 'qrtz_***_triggers', Error_code: 1032; 
14handler error HA_ERR_KEY_NOT_FOUND; the event's master log FIRST, end_log_pos 603058578

由于不同步的时间已经很久了，主库的 binlog 已经不存在，所以只能重建从库了，本文使用 MySQL 8 的 clone 插件进行主从重建。

主从重建

安装插件

主从均需安装 clone 插件：

SQL

1mysql> install plugin clone soname 'mysql_clone.so';
2Query OK
3
4mysql> select plugin_name, plugin_status from information_schema.plugins where plugin_name = 'clone';
5+-------------+---------------+
6| PLUGIN_NAME | PLUGIN_STATUS |
7+-------------+---------------+
8| clone       | ACTIVE        |
9+-------------+---------------+

确保安装完成即可。

配置权限

执行克隆操作需要特定的权限，本地复制和远程复制权限不同：

本地克隆：执行克隆操作的用户需要 BACKUP_ADMIN 权限；
远程克隆：捐赠者（数据源）用户需要 BACKUP_ADMIN 权限；接受者（目标）用户需要 CLONE_ADMIN 权限（该权限隐含了 BACKUP_ADMIN 和 SHUTDOWN 权限）；

主（源端）：

SQL

1mysql> GRANT BACKUP_ADMIN ON *.* TO 'root'@'%';
2mysql> FLUSH PRIVILEGES;

从（目标端）：

SQL

1mysql> GRANT CLONE_ADMIN ON *.* TO 'root'@'%';
2mysql> FLUSH PRIVILEGES;

远程克隆

从库开始克隆：

SQL

1-- 从库配置参数
2mysql> set global clone_valid_donor_list='192.168.31.188:3306';
3
4-- 从库开始克隆
5mysql> clone instance from root@'192.168.31.188':3306 identified by 'P@ssw0rd';

克隆过程中，从库可以查询进度：

SQL

1-- 查看克隆进度
2mysql> SELECT 
3    stage,
4    state,
5    CAST(begin_time AS DATETIME) AS "START TIME",
6    CAST(end_time AS DATETIME) AS "FINISH TIME",
7    LPAD(sys.format_time(POWER(10, 12) * (UNIX_TIMESTAMP(end_time) - UNIX_TIMESTAMP(begin_time))), 10, ' ') AS DURATION,
8    LPAD(CONCAT(FORMAT(ROUND(estimate / 1024 / 1024, 0), 0), "MB"), 16, ' ') AS "Estimate",
9    CASE 
10        WHEN begin_time IS NULL THEN LPAD('%0', 7, ' ')
11        WHEN estimate > 0 THEN LPAD(CONCAT(ROUND(data * 100 / estimate, 0), "%"), 7, ' ')
12        WHEN end_time IS NULL THEN LPAD('0%', 7, ' ')
13        ELSE LPAD('100%', 7, ' ')
14    END AS "Done(%)"
15FROM performance_schema.clone_progress;
16
17+-----------+-----------+---------------------+---------------------+------------+------------------+---------+
18| stage     | state     | START TIME          | FINISH TIME         | DURATION   | Estimate         | Done(%) |
19+-----------+-----------+---------------------+---------------------+------------+------------------+---------+
20| DROP DATA | Completed | 2026-01-05 15:34:10 | 2026-01-05 15:34:13 |     3.07 s |              0MB |    100% |
21| FILE COPY | Completed | 2026-01-05 15:34:13 | 2026-01-05 15:58:13 |       24 m |        347,246MB |    100% |
22| PAGE COPY | Completed | 2026-01-05 15:58:13 | 2026-01-05 15:58:13 |  507.57 ms |             18MB |    100% |
23| REDO COPY | Completed | 2026-01-05 15:58:13 | 2026-01-05 15:58:14 |   368.4 ms |              0MB |    100% |
24| FILE SYNC | Completed | 2026-01-05 15:58:14 | 2026-01-05 15:58:15 |  895.55 ms |              0MB |    100% |
25| RESTART   | Completed | 2026-01-05 15:58:15 | 2026-01-05 15:58:21 |     6.47 s |              0MB |    100% |
26| RECOVERY  | Completed | 2026-01-05 15:58:21 | 2026-01-05 15:58:23 |     1.64 s |              0MB |    100% |
27+-----------+-----------+---------------------+---------------------+------------+------------------+---------+
28
29-- 查看克隆最终状态
30mysql> SELECT * FROM performance_schema.clone_status;
31
32+------+------+-----------+-------------------------+-------------------------+--------------------+----------------+----------+---------------+------------------+-----------------+---------------+
33| ID   | PID  | STATE     | BEGIN_TIME              | END_TIME                | SOURCE             | DESTINATION    | ERROR_NO | ERROR_MESSAGE | BINLOG_FILE      | BINLOG_POSITION | GTID_EXECUTED |
34+------+------+-----------+-------------------------+-------------------------+--------------------+----------------+----------+---------------+------------------+-----------------+---------------+
35|    1 |    0 | Completed | 2026-01-05 15:34:09.386 | 2026-01-05 15:58:22.894 | 192.168.31.188:3306 | LOCAL INSTANCE |        0 |               | mysql-bin.004291 |       630818978 |               |
36+------+------+-----------+-------------------------+-------------------------+--------------------+----------------+----------+---------------+------------------+-----------------+---------------+

等待克隆完成后，会自动重新启动从库 MySQL 实例。

启动主从复制

从库配置并启动主从复制：

SQL

1mysql> show variables like '%server_id%';
2+----------------+-------+
3| Variable_name  | Value |
4+----------------+-------+
5| server_id      | 2     |
6| server_id_bits | 32    |
7+----------------+-------+
8
9-- 配置从库
10mysql> change master to master_host='192.168.31.188',
11master_port=3306,
12master_user='repl',
13master_password='P@ssw0rd',
14master_log_file='mysql-bin.004291',
15master_log_pos=630818978;
16
17-- 启动主从同步
18mysql> start slave;
19Query OK
20
21-- 检查主从同步状态
22mysql> show slave status \G;
23*************************** 1. row ***************************
24               Slave_IO_State: Waiting for source to send event
25                  Master_Host: 192.168.31.188
26                  Master_User: repl
27                  Master_Port: 3306
28                Connect_Retry: 60
29              Master_Log_File: mysql-bin.004291
30          Read_Master_Log_Pos: 633519126
31               Relay_Log_File: mysql-relay-bin.000002
32                Relay_Log_Pos: 2697758
33        Relay_Master_Log_File: mysql-bin.004291
34             Slave_IO_Running: Yes
35            Slave_SQL_Running: Yes
36              Replicate_Do_DB: 
37          Replicate_Ignore_DB: 
38           Replicate_Do_Table: 
39       Replicate_Ignore_Table: 
40      Replicate_Wild_Do_Table: 
41  Replicate_Wild_Ignore_Table: 
42                   Last_Errno: 0
43                   Last_Error: 
44                 Skip_Counter: 0
45          Exec_Master_Log_Pos: 633516410
46              Relay_Log_Space: 2700684
47              Until_Condition: None
48               Until_Log_File: 
49                Until_Log_Pos: 0
50           Master_SSL_Allowed: No
51           Master_SSL_CA_File: 
52           Master_SSL_CA_Path: 
53              Master_SSL_Cert: 
54            Master_SSL_Cipher: 
55               Master_SSL_Key: 
56        Seconds_Behind_Master: 0
57Master_SSL_Verify_Server_Cert: No
58                Last_IO_Errno: 0
59                Last_IO_Error: 
60               Last_SQL_Errno: 0
61               Last_SQL_Error: 
62  Replicate_Ignore_Server_Ids: 
63             Master_Server_Id: 1
64                  Master_UUID: 0df21031-d536-11ee-a0cd-005056a91cf7
65             Master_Info_File: mysql.slave_master_info
66                    SQL_Delay: 0
67          SQL_Remaining_Delay: NULL
68      Slave_SQL_Running_State: Replica has read all relay log; waiting for more updates
69           Master_Retry_Count: 86400
70                  Master_Bind: 
71      Last_IO_Error_Timestamp: 
72     Last_SQL_Error_Timestamp: 
73               Master_SSL_Crl: 
74           Master_SSL_Crlpath: 
75           Retrieved_Gtid_Set: 
76            Executed_Gtid_Set: 
77                Auto_Position: 0
78         Replicate_Rewrite_DB: 
79                 Channel_Name: 
80           Master_TLS_Version: 
81       Master_public_key_path: 
82        Get_master_public_key: 0
83            Network_Namespace: 
841 row in set

可以发现，主从已经恢复正常同步。

写在最后

以前搭建 MySQL 主从都是使用 mysqldump 或者 PXB 备份恢复，耗时耗力。使用 clone 插件，只要符合它的限制条件，将是非常实用便捷的搭建方式。