Re: Critical performance problems on large databases

On Thu, 11 Apr 2002, Nigel J. Andrews wrote:
>
> On Wed, 10 Apr 2002, Tom Lane wrote:
>
> > Gunther Schadow <gunther@aurora.regenstrief.org> writes:
> > > We also noted that a
> > > SELECT COUNT(*) FROM BigQuery;
> > > can take quite a long time and again use a lot of resources,
> > > whereas
> > > SELECT COUNT(smallcolumn) FROM BigQuery;
> > > may be faster and less resource consuming.
> >
> > This is complete nonsense... if anything, the second one will take
> > more cycles, since it has to actually examine a column.
> >
>
> OK, I hate to do this, partly because I seem to remember mention of how this
> sort of thing has to happen like this because the system isn't clever enough to
> do the optimisation, but...
>
>  [sniped]
>

Forgot to say this is on 7.2, I'm currently having some odd difficulty running
the regression tests for 7.2.1


--
Nigel J. Andrews
Director

---
Logictree Systems Limited
Computer Consultants