目錄
- 案例場(chǎng)景
- 解決問題
- 一點(diǎn)總結(jié)
案例場(chǎng)景
今天在線上發(fā)現(xiàn)一個(gè)問題，由于監(jiān)控沒有覆蓋到，某臺(tái)機(jī)器的磁盤被寫滿了，導(dǎo)致線上MySQL主從復(fù)制出現(xiàn)問題。問題如下：
localhost.(none)>show slave status\G
*************************** 1. row ***************************
Slave_IO_State:
Master_Host: 10.xx.xx.xx
Master_User: replica
Master_Port: 5511
Connect_Retry: 60
Master_Log_File:
Read_Master_Log_Pos: 4
Relay_Log_File: relay-bin.001605
Relay_Log_Pos: 9489761
Relay_Master_Log_File:
Slave_IO_Running: No
Slave_SQL_Running: No
Last_Errno: 13121
Last_Error: Relay log read failure: Could not parse relay log event entry.
The possible reasons are: the master's binary log is corrupted (you can check this by running
'mysqlbinlog' on the binary log), the slave's relay log is corrupted (you can check this by
running 'mysqlbinlog' on the relay log), a network problem, the server was unable to fetch a
keyring key required to open an encrypted relay log file, or a bug in the master's or
slave's MySQL code. If you want to check the master's binary log or slave's relay log,
you will be able to know their names by issuing 'SHOW SLAVE STATUS' on this slave.
于是查看error log，發(fā)現(xiàn)error log中的內(nèi)容如下：
2021-03-31T11:34:39.367173+08:00 11 [Warning] [MY-010897] [Repl] Storing MySQL user name or
password information in the master info repository is not secure and is therefore not
recommended. Please consider using the USER and PASSWORD connection options for START SLAVE;
see the 'START SLAVE Syntax' in the MySQL Manual for more information.
2021-03-31T11:34:39.368161+08:00 12 [ERROR] [MY-010596] [Repl] Error reading relay log
event for channel '': binlog truncated in the middle of event; consider out of disk space
2021-03-31T11:34:39.368191+08:00 12 [ERROR] [MY-013121] [Repl] Slave SQL for channel '': Relay
log read failure: Could not parse relay log event entry. The possible reasons are: the master's
binary log is corrupted (you can check this by running 'mysqlbinlog' on the binary log), the
slave's relay log is corrupted (you can check this by running 'mysqlbinlog' on the relay log),
a network problem, the server was unable to fetch a keyring key required to open an encrypted
relay log file, or a bug in the master's or slave's MySQL code. If you want to check the
master's binary log or slave's relay log, you will be able to know their names by issuing 'SHOW
SLAVE STATUS' on this slave. Error_code: MY-013121
2021-03-31T11:34:39.368205+08:00 12 [ERROR] [MY-010586] [Repl] Error running query, slave SQL
thread aborted. Fix the problem, and restart the slave SQL thread with "SLAVE START". We
stopped at log 'mysql-bin.000446' position 9489626
從描述中可以看到，error log是比較智能的，發(fā)現(xiàn)了磁盤問題，并提示我們需要"consider out of disk space"
解決問題
登錄服務(wù)器，很快就發(fā)現(xiàn)是MySQL所在的服務(wù)器磁盤使用率達(dá)到100%了，問題原因跟error log中的內(nèi)容一致。
現(xiàn)在就解決這個(gè)問題。基本的思路就是清理磁盤文件，然后重新搭建復(fù)制關(guān)系，這個(gè)過程似乎比較簡(jiǎn)單，但是實(shí)際操作中，在搭建復(fù)制關(guān)系的時(shí)候出現(xiàn)了下面的報(bào)錯(cuò)：
### 基于gtid的復(fù)制，想重新搭建復(fù)制關(guān)系
localhost.(none)>reset slave;
ERROR 1371 (HY000): Failed purging old relay logs: Failed during log reset
localhost.(none)>reset slave all;
ERROR 1371 (HY000): Failed purging old relay logs: Failed during log reset
第一步：因?yàn)閺?fù)制是基于gtid進(jìn)行的，所以直接記錄show slave status的狀態(tài)后，就可以重新reset slave，并利用change master語(yǔ)句來(lái)重建復(fù)制關(guān)系了。
但是卻出現(xiàn)上面的報(bào)錯(cuò)，從報(bào)錯(cuò)信息看是mysql無(wú)法完成purge relay log的操作，這看起來(lái)不科學(xué)。好吧，既然你自己不能完成purge relay logs的操作，那就讓我來(lái)幫你吧。
第二步：手工rm -f 刪除所有的relay log，發(fā)現(xiàn)報(bào)錯(cuò)變成了：
localhost.(none)>reset slave all;
ERROR 1374 (HY000): I/O error reading log index file
嗯，好吧，問題沒有得到解決。
然后思考了下，既然不能通過手工reset slave 來(lái)清理relay log，直接stop
slave 然后change master行不行呢？
第三步：直接stop slave，然后change master，不執(zhí)行reset slave all的語(yǔ)句，結(jié)果如下：
localhost.(none)>change master to master_host='10.13.224.31',
-> master_user='replica',
-> master_password='eHnNCaQE3ND',
-> master_port=5510,
-> master_auto_position=1;
ERROR 1371 (HY000): Failed purging old relay logs: Failed during log reset
得，問題依舊。
第四步：反正復(fù)制已經(jīng)報(bào)錯(cuò)斷開了，執(zhí)行個(gè)start slave看看，結(jié)果戲劇性的一幕出現(xiàn)了：
localhost.(none)>start slave;
ERROR 2006 (HY000): MySQL server has gone away
No connection. Trying to reconnect...
Connection id: 262
Current database: *** NONE ***
Query OK, 0 rows affected (0.01 sec)
localhost.(none)>
[root@ ~]#
執(zhí)行start slave之后，實(shí)例直接掛了。
到這里，復(fù)制徹底斷開了，從庫(kù)實(shí)例已經(jīng)掛了。
第五步：看看實(shí)例還能不能重啟，嘗試重啟實(shí)例，發(fā)現(xiàn)實(shí)例還能起來(lái)。實(shí)例重新起來(lái)后，查看復(fù)制關(guān)系，結(jié)果如下：
localhost.(none)>show slave status\G
*************************** 1. row ***************************
Slave_IO_State: Queueing master event to the relay log
Master_Host: 10.xx.xx.xx
Master_User: replica
Master_Port: 5511
Connect_Retry: 60
Master_Log_File:
Read_Master_Log_Pos: 4
Relay_Log_File: relay-bin.001605
Relay_Log_Pos: 9489761
Relay_Master_Log_File:
Slave_IO_Running: Yes
Slave_SQL_Running: No
Last_Errno: 13121
Last_Error: Relay log read failure: Could not parse relay log event entry.
The possible reasons are: the master's binary log is corrupted (you can check this by running
'mysqlbinlog' on the binary log), the slave's relay log is corrupted (you can check this by
running 'mysqlbinlog' on the relay log), a network problem, the server was unable to fetch a
keyring key required to open an encrypted relay log file, or a bug in the master's or slave's
MySQL code. If you want to check the master's binary log or slave's relay log, you will be able
to know their names by issuing 'SHOW SLAVE STATUS' on this slave.
Skip_Counter: 0
復(fù)制關(guān)系依舊報(bào)錯(cuò)。
第六步：重新reset slave all看看，結(jié)果成功了。
localhost.(none)>stop slave;
Query OK, 0 rows affected (0.00 sec)
localhost.(none)>reset slave all;
Query OK, 0 rows affected (0.03 sec)
第七步：重新搭建復(fù)制關(guān)系并啟動(dòng)復(fù)制
localhost.(none)>change master to master_host='10.xx.xx.xx',
-> master_user='replica',
-> master_password='xxxxx',
-> master_port=5511,
-> master_auto_position=1;
Query OK, 0 rows affected, 2 warnings (0.01 sec)
localhost.(none)>start slave;
Query OK, 0 rows affected (0.00 sec)
localhost.(none)>show slave status\G
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 10.xx.xx.xx
Master_User: replica
Master_Port: 5511
Connect_Retry: 60
...
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
發(fā)現(xiàn)實(shí)例的復(fù)制關(guān)系可以建立起來(lái)了。
一點(diǎn)總結(jié)
當(dāng)磁盤寫滿的情況發(fā)生之后，mysql服務(wù)無(wú)法向元信息表中寫數(shù)據(jù)，relay log也可能已經(jīng)不完整了，如果直接清理了服務(wù)器上的磁盤數(shù)據(jù)，再去重新change master修改主從復(fù)制關(guān)系，可能會(huì)出現(xiàn)報(bào)錯(cuò)，不能直接修復(fù)，因?yàn)檫@不是一個(gè)正常的主從復(fù)制關(guān)系斷裂場(chǎng)景。
所以，正確的做法應(yīng)該是：
1、清理服務(wù)器的磁盤
2、重啟復(fù)制關(guān)系斷開的那個(gè)從庫(kù)
3、重新reset slave all、change master來(lái)搭建主從復(fù)制關(guān)系即可
如果有更好的方法，還請(qǐng)不吝賜教。
以上就是磁盤寫滿導(dǎo)致MySQL復(fù)制失敗的解決方案的詳細(xì)內(nèi)容，更多關(guān)于MySQL復(fù)制失敗的解決方案的資料請(qǐng)關(guān)注腳本之家其它相關(guān)文章！
您可能感興趣的文章:- MySql主從復(fù)制機(jī)制全面解析
- Mysql主從復(fù)制與讀寫分離圖文詳解
- MySQL 復(fù)制表的方法
- MySQL 8.0.23中復(fù)制架構(gòu)從節(jié)點(diǎn)自動(dòng)故障轉(zhuǎn)移的問題
- MYSQL數(shù)據(jù)庫(kù)GTID實(shí)現(xiàn)主從復(fù)制實(shí)現(xiàn)(超級(jí)方便)
- MySql主從復(fù)制實(shí)現(xiàn)原理及配置
- 淺析MySQL的WriteSet并行復(fù)制
- MySQL主從復(fù)制原理以及需要注意的地方
- mysql 如何動(dòng)態(tài)修改復(fù)制過濾器
- 淺析MySQL并行復(fù)制
- MySQL復(fù)制問題的三個(gè)參數(shù)分析
