On Fri, Nov 20, 2020 at 11:02 AM Alvaro Herrera <alvherre@alvh.no-ip.org> wrote:
> On 2020-Nov-20, Robert Haas wrote:
> > Yeah, I think dirtying the page fewer times is a big win. However, a
> > page may have tuples that are not yet all-visible, and we can't freeze
> > those just because we are freezing others.
>
> Of course! We should only freeze tuples that are freezable. I thought
> that was obvious :-)
I didn't mean to imply that anyone in particular thought the contrary.
It's just an easy mistake to make when thinking about these kinds of
topics. Ask me how I know.
It's also easy to forget that both xmin and xmax can require freezing,
and that the time at which you can do one can be different than the
time at which you can do the other.
Or at least, I have found it so.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company