On Thu, Aug 17, 2006 at 01:47:37PM -0400, Matthew T. O'Connor wrote:
> Alvaro Herrera wrote:
> >Matthew T. O'Connor wrote:
> >
> >>I assume you are suggesting that the base value be 0? Well for one
> >>thing if the table doesn't have any rows that will result in constant
> >>vacuuming of that table, so it needs to be greater than 0. For a small
> >>table, say 100 rows, there usually isn'tn much performance impact if the
> >>table if 50% dead space, so I think the base values you suggest are OK,
> >>but they shouldn't be 0.
> >>
> >
> >Actually Tom suggested some time ago that we should get rid of the base
> >value completely, i.e. make it 0 forever.
> >
> >A row with 0 tables would not show any activity in pgstats, so it would
> >not be vacuumed constantly. Only once after it's truncated.
>
> OK, forgot that. Well I put it in originally as a way to give more
> flexability to the calculation, if I want a tabled vacuumed every 100
> updates, then I can set the scaling factor to 0 and the base value to
> 100, but maybe that's not really needed. It would simplify things if we
> got rid of it.
I think it makes more sense in the per-table settings (which I can't
remember if we actually have yet). For example, on a frequently update
table that you know should always be small, you might well want to set
it to scaling factor 0 and base of 20 or whatever.
--
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