Exec_Master_Log_Pos always behind Read_Master_Log_Pos

MySQL & MariaDB Percona Server for MySQL 5.7
1

I’m checking the health of the replication comparing Exec_Master_Log_Pos and Read_Master_Log_Pos. All is fine in general, except on the one node we have 8.0 instead of 5.7. Exec on that node absolutely never catches up, regardless of load. Hence the health is always shown as degraded.
Seconds_Behind_Master is usually 0, but from time to time I can catch it at 1.

Is this something specific to 8.0, can I do anything about that, should I change the health check system, is it healthy in the end?

              Master_Log_File: mysql-bin.002557
          Read_Master_Log_Pos: 183729631
               Relay_Log_File: srv-relay-bin.001184
                Relay_Log_Pos: 183727785
        Relay_Master_Log_File: mysql-bin.002557
             Slave_IO_Running: Yes
            Slave_SQL_Running: Yes
              Replicate_Do_DB:
          Replicate_Ignore_DB:
           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: 183727569
              Relay_Log_Space: 183730102
2

Hello @Add,
Be sure to compare all global variables between 5.7 and 8.0 servers. Make sure you are using ROW-based replication. Other than that, check out the PMM dashboard for the Performance Schema on your 8.0 and see where you might be having wait issues.

3

These are the differences (usually the defaults between those 2 versions):

Replication is ROW based, I haven’t PMM installed.