Re: reloption to prevent VACUUM from truncating empty pages at the end of relation

Andres Freund wrote:
> On 2018-04-19 16:56:59 -0300, Alvaro Herrera wrote:
> > Michael Paquier wrote:
> >
> > > Then, let's consider the beginning of the first commit fest of v12 as
> > > judgement. Implementing radix tree for shared buffers is a long-term
> > > project, which has no guarantee to get merged, while a visibly-simple
> > > reloptions which helps in some cases...
> >
> > In the scenario we studied, the truncations were causing periodic
> > hiccups which were quite severe.
>
> Was that with the current logic of breaking the truncations into smaller
> chunks?

Yes -- it was with 9.5.7. I was skeptical about that stuff working
correctly for a toast table, BTW, but I didn't manage to prove anything.

> > The truncations were completely useless anyway because the table
> > grew back to the original size daily (a few dozen GBs I think).
> > That was a lot of unnecessary work, and under exclusive lock no
> > less.
>
> FWIW, One goal of the different buffer mapping implementation is to also
> make both increasing and decreasing size of relations possible without
> an AEL.

Oh, that sounds very useful.

--
Álvaro Herrera https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services