The processing process of MySQL master-slave synchronization fault

Today there are operations of the classmate response, a business master-slave Replication (a main two from) a large delay;

MySQL master:5.0.82 slave:5.0.88

First point out that the DB data volume or dozens of g size, the network is not a problem, OS system load is small, Io is not tight;

Phenomenon 1, slave A appears thousands of relay_log, and the size of a few KB, error log no other error display;

The phenomenon 2, slave B show slave status\g display Normal, but the error log generates the wrong file on G, the fault message is slave IO error can not read Master Binlog;

Reboot slave B cannot be resolved;

Stop slave a copy, SLAVEB not generate error log, aserver error message is recorded on B server;

The above phenomenon can not find clues, not to mention to monitor whether the lock problem causes the master-slave replication delay is big;

The only thing you can look at is the server_id of slave A, slave b;

The server_id of both are the same;;;

The DB installation portion of the company is also transferred from the operational dimension to the DBA side.

This article is from the "Technology Achievement Dream" blog, please be sure to keep this source http://weipengfei.blog.51cto.com/1511707/1208792

See more highlights of this column: http://www.bianceng.cnhttp://www.bianceng.cn/database/MySQL/