The value of Seconds_Behind_Master is increasing

keith · December 30, 2020, 4:24pm

I saw on the slave Seconds_Behind_Master increasing.But the master did not perform large transactions.

This is ‘show slave status \G’ part of the results：
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
Replicate_Do_DB:
Replicate_Ignore_DB: test
Replicate_Do_Table:
Replicate_Ignore_Table:
Replicate_Wild_Do_Table:
Replicate_Wild_Ignore_Table:
Last_Errno: 0
Last_Error:
Skip_Counter: 0
Exec_Master_Log_Pos: 164396778
Relay_Log_Space: 654817698
Until_Condition: None
Until_Log_File:
Until_Log_Pos: 0
Master_SSL_Allowed: No
Master_SSL_CA_File:
Master_SSL_CA_Path:
Master_SSL_Cert:
Master_SSL_Cipher:
Master_SSL_Key:
Seconds_Behind_Master: 10995
Master_SSL_Verify_Server_Cert: No
Last_IO_Errno: 0
Last_IO_Error:
Last_SQL_Errno: 0
Last_SQL_Error:
Replicate_Ignore_Server_Ids:
Master_Server_Id: 193
SQL_Delay: 0

when I run ‘show full processlist’ I can see UPDATE ,and the id is constantly changing.There is no data written in addition to master,but the value of Seconds_Behind_Master is increasing.

Master percona server version is 5.6.36-82.0-log,system version is centos7.
Slave percona server version is 5.6.36-82.1-log,system version is centos5.

Tell me the way to check,Please.
Thanks.

nvngpt31 · December 30, 2020, 4:34pm

Can you confirm like Exec_Master_Log_Pos: is changing?

keith · December 30, 2020, 4:34pm

Thank you for helping me .
Yes , It is changing. I found the reason for the problem. Scheduled tasks on the master server is working without index. And the disk of master server is SSD , the slave is HHD. Each modification of a record requires a full table scan.I stopped the scheduled task and returned to normal.