One reason for slow use of not in MySQL is that the field type does not match.

Recently, the customer's response to a page is getting slower and slower. It takes more than a dozen seconds to complete the operation.

SQL is not complex, probably because: Select field from table 1 where field 1 not in (select Field 2 from table 2)

The number of tables 1 and 2 is tens of thousands, and the data volume is not large. The technicians checked the index and did not change the data with not exists and left join.

It took several days to find that the field type of field 1 is int, and the field type of field 2 is varchar. therefore, the field type of field 2 is changed to int, and the result execution speed is within 1 second.