On Fri, Aug 04, 2006 at 02:40:30PM -0400, Tom Lane wrote:
> which it seems we ought to be bright enough to notice. In particular
> I would argue that turning on constraint_exclusion ought to instruct
> the planner to catch this sort of thing, whereas when it's off we
> ought not expend the cycles. I have a preliminary patch (below)
> that seems to fix it.
How many cycles are we talking about here? Is it even worth the GUC?
The most heavily loaded systems I've seen were doing on the order of 500
transactions a second (which would have been almost entirely
single-statement queries), so ISTM that you've got to be burning a
pretty good chunk of CPU time on planning before it becomes an issue.
--
Jim C. Nasby, Sr. Engineering Consultant jnasby@pervasive.com
Pervasive Software http://pervasive.com work: 512-231-6117
vcard: http://jim.nasby.net/pervasive.vcf cell: 512-569-9461