server_id conflict causes IO wait failure

Problem Description:

After adding a new MySQL Slave on the line, the server is abnormal.

1, show processlist;

Queueing master event to the relay logreconnecting after a failed master event read

# repeated execution of the instruction, found that slave is constantly trying to register to master, but registration failed.
# Execute show slave status\g find master-slave is in sync state.

2, the server disk IO occupation is more serious

# Top Display 1.9%wa
# iotop Display 9.8M WRITE per second

Workaround:

First think of the operation, the new addition of Slave, the way is to copy the data directory from another Slave directly with the configuration file way.

Login This Slave found the same as the new Slave.

Recall just the operation, found that the new Slave forgot to modify the server_id, started the service.

Change the new Slave server_id to globally unique, restart MySQL and find everything back to normal.

server_id conflict causes IO wait failure