With or without DISTINCT ???

Question

With or without DISTINCT ???

T.Ashish

SSCarpal Tunnel

Points: 4248
More actions
April 26, 2013 at 1:40 am

#277196

Hi all,
Here are two very small queries in which only difference is the use of DISTINCT. Both the queries are returning same rows and have identical IO Statistics and execution plan.
So, technically as there there is no performance gain we can use either of them. But I still want to use 2nd query as it is not having DISTINCT clause.
Can you share your experience on this.
1.
SELECT
top 100000 *
FROM
account_site
WHERE
account_id IN (
SELECT
DISTINCT account_id
FROM
account
)
2.
SELECT
top 100000 *
FROM
account_site
WHERE
account_id IN (
SELECT
account_id
FROM
account
)
Execution plan attached.

Viewing 8 posts - 1 through 7 (of 7 total)

You must be logged in to reply to this topic. Login to reply

John Mitchell-245523 SSC Guru Points: 148809 More actions · Answer 1

Is there a PRIMARY KEY or UNIQUE constraint on account.account_id? If there is, that'll be why the execution plans are identical.

John

T.Ashish SSCarpal Tunnel Points: 4248 More actions · Answer 2

T.Ashish

SSCarpal Tunnel

Points: 4248

April 26, 2013 at 3:07 am

#1610252

Yes, There is a primary key on account.account_id

Michael L John One Orange Chip Points: 27038 More actions · Answer 3

That's your answer. Account_id is the primary key. It is already unique. The use of distinct will produce the same number of rows as the query without it.

If you are trying to make performance better, try using EXISTS or a simple inner join.

Michael L John
If you assassinate a DBA, would you pull a trigger?
To properly post on a forum:
http://www.sqlservercentral.com/articles/61537/

Kingston Dhasian SSCoach Points: 19799 More actions · Answer 4

Michael L John (4/26/2013)
That's your answer. Account_id is the primary key. It is already unique. The use of distinct will produce the same number of rows as the query without it.
If you are trying to make performance better, try using EXISTS or a simple inner join.

Using INNER JOIN may change the results as the query here is just performing an existence check.

EXISTS might give a different plan and probably better performance with the same output.

Kingston Dhasian

How to post data/code on a forum to get the best help - Jeff Moden
http://www.sqlservercentral.com/articles/Best+Practices/61537/

T.Ashish SSCarpal Tunnel Points: 4248 More actions · Answer 5

I have tested same scenario with a non unique and unindexed column. Results are anyways same.

Yes, I have many other ways to perform the task. But point here is to test the query with or without DISTINCT.

In theory DISTINCT should not be used, but if it is giving same results then why/why not use DISTINCT!!!

John Mitchell-245523 SSC Guru Points: 148809 More actions · Answer 6

Use DISTINCT when you want to eliminate duplicates from your result test. If you use it in an existence test, like you did in your original post, then it won't make any difference to the results. That applies whether the column has a unique/PK constraint or not.

John

ChrisM@Work SSC Guru Points: 186127 More actions · Answer 7

John Mitchell-245523 (4/26/2013)
Use DISTINCT when you want to eliminate duplicates from your result test. If you use it in an existence test, like you did in your original post, then it won't make any difference to the results. That applies whether the column has a unique/PK constraint or not.
John

+1

DISTINCT used in the EXISTS check is logically irrelevant. Check the plans - it may even be optimized away.

^{“Write the query the simplest way. If through testing it becomes clear that the performance is inadequate, consider alternative query forms.” - Gail Shaw}

For fast, accurate and documented assistance in answering your questions, please read this article.
Understanding and using APPLY, (I) and (II) Paul White
Hidden RBAR: Triangular Joins / The "Numbers" or "Tally" Table: What it is and how it replaces a loop Jeff Moden