I confirmed this with a table I have with 9 million rows.  Count(*)
always returned in time <= count(field).  But, as Adam said, watch your
indexed columns in your where clause.  But Count(*) should still be
faster.


-----Original Message-----
From: Raster, Tim [mailto:[EMAIL PROTECTED]
Sent: Tuesday, July 13, 2004 13:32
To: SQL
Subject: RE: Faster count?


Actually, I think count(*) is faster.  It is a special signal to the
server that it doesn't have to count field values, but rows, which saves
time having to evaluate rows.  Unlike select *, it does NOT mean all
fields... it means all rows.  So count(*) is faster.

-----Original Message-----
From: Adam Haskell [mailto:[EMAIL PROTECTED]
Sent: Tuesday, July 13, 2004 13:28
To: SQL
Subject: Re: Faster count?

no...its slower. Always count an indexed column if possible.

----- Original Message -----
From: Michael Dinowitz <[EMAIL PROTECTED]>
Date: Tue, 13 Jul 2004 14:24:59 -0400
Subject: Faster count?
To: SQL <[EMAIL PROTECTED]>

Is count (*) faster than count(specific_field)?
Thanks
--
Michael Dinowitz
http://www.houseoffusion.com
For all your ColdFusion needs________________________________
________________________________
________________________________
[Todays Threads] [This Message] [Subscription] [Fast Unsubscribe] [User Settings] [Donations and Support]

Reply via email to