Comments on: SQL SERVER – Optimization Rules of Thumb – Best Practices – Reader’s Article

By: Matt Chatterley

Matt Chatterley — Tue, 23 Dec 2008 14:08:56 +0000

Hi,

Some good pointers in the above to get started with – to answer previous question, exists rather than distinct:

SELECT Column
FROM dbo.Table1 T1
WHERE EXISTS
( SELECT 1 FROM dbo.Table 2 T2 WHERE T2.Column=T1.Column )

Instead of:

SELECT DISTINCT Column FROM dbo.Table1

Assuming a parent/child relationship between T1 and T2 – this doesn’t apply to all cases, of course. If you have duplication in tables which are not part of a parent/child relationship, it is likely that you need to normalize the data further!

By: Kamran Shahid

Kamran Shahid — Mon, 28 Jul 2008 06:38:53 +0000

How To get distinct record by using Exists rather then Distinct
["where possible use EXISTS rather than DISTINCT"]

By: tong

tong — Mon, 30 Jun 2008 19:05:59 +0000

how to write query can minimize the number of table lookups especially if there are sub-query SELECTs or multicolumn UPDATEs.

thanks,
tong

By: Lisber

Lisber — Fri, 09 May 2008 10:56:28 +0000

Hello, this is my first comment on this excellent Blog !!!.

One question that i’ve been asking myself lately and related with sql server query performance is:

How can i get only the registries between a certain interval, for example in Mysql i use ‘LIMIT Start , END’ in order to limitate the returning set of values.

How can i achive a similar behavior on SQL Server 2005 ??

This tip can dramatically improve many aplications that i’m developing right now, thank and keep the good work.

By: Rahul Sharma

Rahul Sharma — Mon, 05 May 2008 06:23:11 +0000

Nice Explanation