I think you need non-clustered index on the following fields:

PublishDate

SectionID

Headline

Please correct me Pinal if I am wrong.
Thanks.

I was given database. They added some tables. And I need to improve database optimization of table. What is the best way to start and what other things it requires?

Thanks,
Krunal

I going through the implementation of indexes on following given table and will use following reterival queries to retreive the data. Can you please help me out which column(s) I should use for Clustered index & for Non Clustered and for Included Columns index for covered queries.. with example..

–Table Defination starts–
Create Table tblNews
(
SectionID int,
NewsID int,
Headline varchar(500),
Short_Story varchar(500),
PublishDate DateTime,
CONSTRAINT pk_NewsID Primary Key (NewsID)
)
–Table Defination ends–

–Queries—

select Headline, Short_Story from tblNews order by PublishDate desc

select Headline, Short_Story from tblNews where SectionID=1 order by PublishDate desc

select Headline, Short_Story from tblNews where NewsID between 1000 and 1500

select Headline, Short_Story from tblNews where Headline like ‘%USA%’order by PublishDate desc

—-
Please do suggest and it will be a great help.

Thanks
Gaurav Kukar

I am a developer from pvt. firm in navi mumbai. I am developing an application which will handle entire email server transactions by an IT person. To develop this application using .net 2008 and Sql server 2005, i have 2 tables having records 2,500,000+ and 150,000+ respectively in its initial stage which will grow to 10 folds possibly in future. I want to have search queries on these tables to show updates through this application. WIth this table i have 3 more tables having records less than 35000 each which will be used with inner joins to display specific info on application.

Now , i want your help on how should i design queries or indexes which will fetch me desired results without giving connection timeout error. And also it should not take considerable time to load.My seniors have challenged me to fix this issue. I dont want to keep anything for them from my side. So please help me on this.

Please reply me on above email id by or before monday so i shold progress asap on this development.

What do you mean by Not working.

What is not working. Please provide more details.

~ IM.

i have converted non clustered index to clustered index
by using the below query

create unique clustered index PK_Activity on Activity(Activity_id) WITH DROP_EXISTING

some times it is not working because the table have
foreign key

how to avoid the foreign key problem

You can try one of these,

1. SP_depends [object_name]

2. SP_help [object_name]

3. From SQL Server Management Studio, in Object Explorer, right click Table Name – View dependencies.

~ IM.

- a poorly-chosen clustered index key can provoke table scans where an effective one will enable seeks, especially in the case of range queries.
- a poor clustered index will definitely result in a high rate of table and index fragmentation, as alluded to by Marco C above.
- the clustered index key is implicitly included in all non-clustered indices, so choosing one that is too wide will slow performance and increase the size of all non-clustered indices on the same table.

So saying things like “Clustered Index improves performance,” as you do in the comments, is a bad idea. I urge you to review your understanding of SQL Server data structures so that you do not promote potentially damaging ideas in your role as a Microsoft MVP.

Follow this link, there is a already written script which you can refer to, this script will create a script for all indexes in the database, including primary key and unique key indexes.

http://www.sqlservercentral.com/scripts/Index+Management/31652/

If link does not works, do this,

go to google.com, type “how to script all indexes in a database sql server” and click the first link.

Hope this helps.
Imran.

Could you please tell me the scenario in which i should opt for custered index or non-clustered index. What is the trade off between two?

I want to know the F4 (search for objects) procedures in 2005. How to do that?

Thaks