Groups | Blog | Home
all groups > sql server programming > august 2006 >

sql server programming : Ntext optimization!


SqlBeginner
8/14/2006 8:12:02 PM
Hi All,

I have created a DB design for implementing WIKI in one of my project. I
have created a column of data type NText to store the content of the wiki
page. Normally ppl would store the wiki pages as flat files but since my
project work on a web farm scenario I need to save the contents within a DB
(as per our company policy... lets not get deep into this :) ).

Though my app is working fine. When I try to search for a document at
frequent intervals its throwing a "timeout error". [I don't think that
increasing the timeout minutes in the front end would solve this on a long
term basis]. I guess its because of this large data type. FYI, as of now we
have just few hundreds of rows.

Any pointers on how to avoid this error would be really helpful to me.

Regards
Uri Dimant
8/15/2006 12:00:00 AM
Hi
Have you considered using Full-Text Catalog for searching ?


[quoted text, click to view]

Dan Guzman
8/15/2006 12:00:00 AM
I can see that queries will be slow if the content is large and you are
searching the column using LIKE because that will require a table scan as
well as reading all ntext data in the table. Uri's full-text suggestion is
probably the way to go. Another option is to create a keyword table that
relates pages/keywords. The keyword table method is very fast but requires
that you associate keywords to pages, either manually or with automated
document parsing. FT does this automatically and provides more robust
search features.

--
Hope this helps.

Dan Guzman
SQL Server MVP

[quoted text, click to view]

Uri Dimant
8/16/2006 12:00:00 AM
Hi
BOL is the palce to start with.
Also , there is a dedicated sql server forum for full-text search questions
like this one





[quoted text, click to view]

SqlBeginner
8/16/2006 1:22:01 AM
Sorry for responding late. Can you educate me on how to enable or make use of
Full text search in sql server 2000?

Any pointers regarding this would be really helpful.

Regards
Pradeep

[quoted text, click to view]
AddThis Social Bookmark Button