Question

So, I’ve been seeing a lot of SQL examples on this site. I have a question about the relative performance of inner joins (plain JOIN) and cross joins (SELECT foo FROM bar,baz WHERE). Turns out the question has already been asked:

INNER JOIN ON vs WHERE clause

But I still have an issue I’d like clarification on. I didn’t see anything in the answers.

The questions is this:

Assume no fields are NULL. Given two equivalent queries, one formulated like this:

SELECT * FROM t1
JOIN t2 ON t1.t2_id=t2.t1_id AND t2.bar='baz'
WHERE t1.foo='bar'

And one formatted like this:

SELECT * FROM t1,t2
WHERE t1.foo='bar' AND t1.t2_id=t2.t1_id AND t2.bar='baz'

Is there a difference in their execution time? I’m interested specifically in the case where restrictions are placed on values located in both tables, in addition to the ID-matching to associate like rows. Note that there is no foreign key constraint in this schema.

I should probably also say that I’m interested in how this extends to more than two tables.

Thanks in advance for your answers, SQL experts!

Answer 1

Your first example is normally called an explicit join and the second one an implicit join. Performance-wise, they should be equivalent, at least in the popular DBMSes.