I seem to remember that Microsoft SQL Server has the concept of a "covered index" that
worked something like this:
Suppose you have a table with a product_category, and a price. You query alot on
product_category, so you create an index on it. You're expression might look like
this:
select product_id, product_category, price from product where product_category=15245
In SQL Server, you could optimize this by creating your index using all three fields.
The MySQL equivalent would be
create index whatever (product_category, product_price, product_id)
Now, using the same select statement as above, the server would return all data
strictly from the index. It wouldn't even read from the table.
Does MySQL do this now, or are there plans to do something similar in the future?
