On Thu, Jan 27, 2005 at 21:17:56 -0800,
Zavier Sheran <zsheran@yahoo.com> wrote:
> quote from manual:
> --
> Unfortunately, there is no similarly trivial query
> that can be used to improve the performance of count()
> when applied to the entire table
> --
>
> does count(1) also cause a sequential scan of the
> entire table? It should be able to just use the
> primary keys.
No it can't just use the index file, so that an index scan will be slower
than the sequential scan unless there is a where clause restricting the
number of rows to a small fraction (about 5%) of the table.
Search the archives for if you want to read more about this.