Re: scanning 2 rows slow index fast 26GB MyISAM

[Jeff Smelser]

On Thursday 23 December 2004 02:01 pm, Bryan Heitman wrote:
> Hi Sergio,
>
> All of your suggestions deal with key optimization, I do not believe I have
> a key issue here.  Remember that select count(*), an index-only query
> returns in .06 seconds which is very quick.  The real question, is why does
> it take 5 mins to retrieve the row data for these 2 rows that the index
> retrieved so quickly.  Why the delay and why the heavy read activity on the
> MYD file.

Do a show table status. It shows the number of rows in a table. It knows this 
well before you run select count(*), so it really doesnt count anything. 

So the select count(*) theory isnt correct.

Jeff

Attachment: [application/pgp-signature]