Question

If I issue the following commands:

CREATE TABLE foo LIKE bar;

INSERT INTO bar SELECT * FROM foo;

There are several million rows in foo, so this action takes some time.

If I do the following query repeatedly during the execution of the above:

use information_schema
select table_name,  TABLE_ROWS  from tables where table_name IN ('foo','bar');

I’ve seen the following progression:

+---------------------------------+------------+
| table_name                      | TABLE_ROWS |
+---------------------------------+------------+
| foo                             |   21270328 |
| bar                             |    5292482 |
+---------------------------------+------------+

and then a little later:

+---------------------------------+------------+
| table_name                      | TABLE_ROWS |
+---------------------------------+------------+
| foo                             |   21270328 |
| bar                             |    3422182 |
+---------------------------------+------------+

Why does this number decrease?

Answer 1

As noted in the documentation for information_schema.tables, “For InnoDB tables, the row count is only a rough estimate used in SQL optimization.”