Re: slow vacuum performance

On Wed, 24 Mar 2004, pginfo wrote:

> Hi,
>
> scott.marlowe wrote:
>
> > On Wed, 24 Mar 2004, pginfo wrote:
> >
> > > Hi,
> > >
> > > I am running pg 7.4.1 on linux box.
> > > I have a midle size DB with many updates and after it I try to run
> > > vacuum full analyze.
> >
> > Is there a reason to not use just regular vacuum / analyze (i.e. NOT
> > full)?
> >
>
> Yes, in case I make massive updates (only in my case of cource)   for example
> 2 M rows, I do not expect to have 2M new rows in next 180 days.That is the
> reaso for running vacuum full.
> My idea was to free unneedet space and so to have faster system.
> It is possible that I am wrong.

It's all about percentages.  If you've got an average of 5% dead tuples
with regular vacuuming, then full vacuums won't gain you much, if
anything.  If you've got 20 dead tuples for each live one, then a full
vacuum is pretty much a necessity.  The generally accepted best
performance comes with 5 to 50% or so dead tuples.  Keep in mind, having a
few dead tuples is actually a good thing, as your database won't grow then
srhink the file all the time, but keep it in a steady state size wise.