Debian Bullseye下MariaDB频繁重启求助（含InnoDB致命错误）

问题背景 #

服务器经Debian Stretch→Buster→Bullseye三次升级后，MariaDB（版本10.5.15-MariaDB-0+deb11u1-log）出现频繁重启现象（每小时6-10次），导致邮件服务因无法访问虚拟表而中断。

关键错误日志 #

系统服务日志 #

执行命令查看服务状态日志：

cat /var/log/syslog | grep "mariadb.service"

返回内容：

Aug 19 10:34:45 srv systemd[1]: mariadb.service: Main process exited, code=killed, status=6/ABRT
Aug 19 10:34:45 srv systemd[1]: mariadb.service: Failed with result 'signal'.
Aug 19 10:34:50 srv systemd[1]: mariadb.service: Scheduled restart job, restart counter is at 401.

InnoDB致命错误日志 #

Aug 19 10:34:45 srv mariadbd[22091]: 2022-08-19 10:34:45 11917 [ERROR] [FATAL] InnoDB: Page old data size 15870 new data size 8280, page old max ins size 36 new max ins size 7626
Aug 19 10:34:45 srv mariadbd[22091]: 220819 10:34:45 [ERROR] mysqld got signal 6 ;
Aug 19 10:34:45 srv mariadbd[22091]: This could be because you hit a bug. It is also possible that this binary
Aug 19 10:34:45 srv mariadbd[22091]: or one of the libraries it was linked against is corrupt, improperly built,
Aug 19 10:34:45 srv mariadbd[22091]: or misconfigured. This error can also be caused by malfunctioning hardware.
Aug 19 10:34:45 srv mariadbd[22091]: We will try our best to scrape up some info that will hopefully help
Aug 19 10:34:45 srv mariadbd[22091]: diagnose the problem, but since we have already crashed,
Aug 19 10:34:45 srv mariadbd[22091]: something is definitely wrong and this may fail.
Aug 19 10:34:45 srv mariadbd[22091]: Server version: 10.5.15-MariaDB-0+deb11u1-log
Aug 19 10:34:45 srv mariadbd[22091]: key_buffer_size=792723456
Aug 19 10:34:45 srv mariadbd[22091]: read_buffer_size=131072
Aug 19 10:34:45 srv mariadbd[22091]: max_used_connections=15
Aug 19 10:34:45 srv mariadbd[22091]: max_threads=2002
Aug 19 10:34:45 srv mariadbd[22091]: thread_count=12

翻译后核心信息：

2022-08-19 10:34:45 11917 [错误] [致命] InnoDB：页面旧数据大小15870，新数据大小8280；页面旧最大插入大小36，新最大插入大小7626
220819 10:34:45 [错误] mysqld收到信号6；
这可能是触发了程序bug，也可能是二进制文件或依赖库损坏、编译不当、配置错误，或是硬件故障导致。
我们会尽力收集诊断信息，但由于服务已崩溃，部分信息可能无法获取。
服务器版本：10.5.15-MariaDB-0+deb11u1-log
key_buffer_size=792723456
read_buffer_size=131072
max_used_connections=15
max_threads=2002
thread_count=12

解决建议 #

优先尝试InnoDB修复，暂不直接重装 #

1. 强制InnoDB恢复

   • 编辑MariaDB配置文件（通常为/etc/mysql/mariadb.conf.d/50-server.cnf），在[mysqld]段落添加：

     innodb_force_recovery = 1
     

   • 重启服务：systemctl restart mariadb
   • 若仍崩溃，逐步将恢复级别提高到2、3、4（注意：级别4及以上可能导致数据丢失，仅作为最后尝试）
   • 服务启动后立即全量备份数据库：mysqldump --all-databases > all_databases_backup.sql

2. 检查文件系统与硬件

   • 运行fsck检查磁盘文件系统完整性（需先卸载对应分区或进入单用户模式）
   • 用smartctl检查磁盘健康：smartctl -H /dev/sda（替换为实际磁盘设备）
   • 检测内存错误：运行memtester或mcelog工具

3. 验证升级后的依赖与配置

   • 检查依赖库是否完整：ldd /usr/sbin/mariadbd | grep -i "not found"
   • 恢复默认配置对比测试：

     mv /etc/mysql/mariadb.conf.d/50-server.cnf /etc/mysql/mariadb.conf.d/50-server.cnf.bak
     cp /usr/share/mysql/mariadb.conf.d/50-server.cnf /etc/mysql/mariadb.conf.d/
     

   • 重启服务观察是否恢复正常

最终选项：彻底卸载重装（仅修复无效时使用） #

1. 备份所有数据

   • 确保已完成全量数据库备份，同时备份配置文件：cp -r /etc/mysql/ /root/mysql_config_backup/

2. 彻底清理相关包与文件

   systemctl stop mariadb
   apt purge mariadb-server mariadb-client mariadb-common mysql-common
   apt autoremove --purge
   rm -rf /var/lib/mysql/ /var/log/mysql/ /etc/mysql/
   

3. 重新安装MariaDB

   apt update
   apt install mariadb-server mariadb-client
   

4. 恢复数据与配置

   • 导入数据库备份：mysql < all_databases_backup.sql
   • 逐步恢复自定义配置（避免直接覆盖默认配置，防止再次引入问题）

内容的提问来源于stack exchange，提问作者Kenny