MySQL移动数据目录后启动失败的解决方法

背景概述

由于安装数据库时将MySQL的数据目录放在了根目录下，现在存储空间不足，想通过mv将数据目录移动到其他目录下，但将数据目录移动到其他数据目录后，启动数据库失败。

问题复现

本次测试基于 MySQL 8.0.31

1.关闭数据库

mysql> shutdown;
Query OK, 0 rows affected (0.02 sec)

2.查看当前数据目录所在位置

shell> pwd
/mysql80

3.通过mv移动整个MySQL数据目录到其他目录

shell>  mv /mysql80 /data
shell>  cd /data/mysql80/svr
shell>  ln -s mysql-8.0.31-linux-glibc2.12-x86_64 mysql

4.修改属主属组

shell> chown -R mysql.mysql /data

5.修改配置文件中数据目录的地址

shell> sed -i 's#/mysql80#/data/mysql80#g' my5001.cnf

6.启动数据库

shell> /data/mysql80/svr/mysql/bin/mysqld_safe \
--defaults-file=/data/mysql80/conf/my5001.cnf --user=mysql &

此时启动数据库失败，错误日志报错如下：

mysqld: File '/mysql80/dbdata/data5001/log/binlog.000012' not found (OS errno 2 - No such file or directory)
2023-02-27T10:38:09.240576+08:00 0 [ERROR] [MY-010958] [Server] Could not open log file.
2023-02-27T10:38:09.240657+08:00 0 [ERROR] [MY-010041] [Server] Can't init tc log
2023-02-27T10:38:09.240718+08:00 0 [ERROR] [MY-010119] [Server] Aborting
2023-02-27T10:38:10.548605+08:00 0 [System] [MY-010910] [Server] /data/mysql80/svr/mysql/bin/mysqld: Shutdown complete (mysqld 8.0.31)  MySQL Community Server - GPL.

这里报错显示找不到binlog文件，并且报错显示的binlog的目录还是之前的，但是配置文件中的目录已经修改

shell> grep 'log-bin' my5001.cnf 
log-bin=/data/mysql80/dbdata/data5001/log/binlog
log-bin-trust-function-creators

7.问题解决

最后通过查找发现 binlog.index 文件中存放着每个binlog文件的绝对路径地址，这里的路径还是之前路径，内容如下：

shell> cat binlog.index 
/mysql80/dbdata/data5001/log/binlog.000001
/mysql80/dbdata/data5001/log/binlog.000002
/mysql80/dbdata/data5001/log/binlog.000003
/mysql80/dbdata/data5001/log/binlog.000004
/mysql80/dbdata/data5001/log/binlog.000005
/mysql80/dbdata/data5001/log/binlog.000006
/mysql80/dbdata/data5001/log/binlog.000007
/mysql80/dbdata/data5001/log/binlog.000008
/mysql80/dbdata/data5001/log/binlog.000009
/mysql80/dbdata/data5001/log/binlog.000010
/mysql80/dbdata/data5001/log/binlog.000011
/mysql80/dbdata/data5001/log/binlog.000012

修改binlog.index文件中binlog的绝对路径：

shell> sed -i 's#/mysql80#/data/mysql80#g' binlog.index
shell> cat binlog.index 
/data/mysql80/dbdata/data5001/log/binlog.000001
/data/mysql80/dbdata/data5001/log/binlog.000002
/data/mysql80/dbdata/data5001/log/binlog.000003
/data/mysql80/dbdata/data5001/log/binlog.000004
/data/mysql80/dbdata/data5001/log/binlog.000005
/data/mysql80/dbdata/data5001/log/binlog.000006
/data/mysql80/dbdata/data5001/log/binlog.000007
/data/mysql80/dbdata/data5001/log/binlog.000008
/data/mysql80/dbdata/data5001/log/binlog.000009
/data/mysql80/dbdata/data5001/log/binlog.000010
/data/mysql80/dbdata/data5001/log/binlog.000011
/data/mysql80/dbdata/data5001/log/binlog.000012

8.启动数据库

shell> /data/mysql80/svr/mysql/bin/mysqld_safe 
--defaults-file=/data/mysql80/conf/my5001.cnf --user=mysql &

数据库启动成功。

9.作为从节点

需要注意的是，如果该实例还作为其他实例的从节点，还需要设置 relaylog.index 文件中relay log的绝对路径，否则会报如下错误： 错误日志报错：

2023-02-27T15:56:55.224372+08:00 0 [ERROR] [MY-010599] [Repl] log /mysql80/dbdata/data5002/log/relaylog.000002 listed in the index, but failed to stat.
2023-02-27T15:56:55.224422+08:00 0 [ERROR] [MY-011059] [Repl] Error counting relay log space.
2023-02-27T15:56:55.226571+08:00 0 [ERROR] [MY-010426] [Repl] Slave: Failed to initialize the master info structure for channel ''; its record may still be present in 'mysql.slave_master_info' table, consider deleting it.
2023-02-27T15:56:55.226622+08:00 0 [ERROR] [MY-010529] [Repl] Failed to create or recover replication info repositories.

执行 start replica 时也会报错：

# 客户端报错
mysql> start replica;
ERROR 1872 (HY000): Slave failed to initialize relay log info structure from the repository
# error log报错
2023-02-27T15:57:53.858798+08:00 8 [ERROR] [MY-013124] [Repl] Slave SQL for channel '': Slave failed to initialize relay log info structure from the repository, Error_code: MY-013124

10.问题解决

修改 relaylog.index 文件中relay log的绝对路径

sed -i 's#/mysql80#/data/mysql80#g' relaylog.index

重新启动数据库，并启动主从复制

# 重启实例
mysql> restart;
# 启动主从复制
mysql> start replica;

此时主从复制恢复正常。