Thanks for the insight Rod.
is there any other place i can know more abt these principles?
But if the table is clustered then the pages are stored catagiously
with respect to that column rite?
> On Fri, 2002-11-01 at 08:03, mallah@trade-india.com wrote:
>>
>> Hi Rod ,
>>
>> Does it means that index scan is used for less frequenlty occuring data? yes my table was not
>> clustered.
>>
>> can u tell me what does 0.00..6788.24 and rows and width means?
>>
>> in explain out put cost=0.00..6788.24 rows=30001 width=4
>>
>>
>> I have one more table where i face the similar problem , i have not dump - reloaded it yet ,
>> will post again if i face the problem.
>
> Keep in mind that an index scan is very expensive in regards to a single tuple. It has to run
> through (fetch) the index pages, then fetch the pages from the table. Since the table fetches
> are random, the harddrive will probably incur a seek for each tuple found in the index. The
> seeks add up much quicker than a sequential scan (without nearly as many seeks or drive head
> movements).
>
> --
> Rod Taylor
-----------------------------------------
Get your free web based email at trade-india.com.
"India's Leading B2B eMarketplace.!"
http://www.trade-india.com/