Re: VACUUM FULL vs dropped columns

From: Andrew Dunstan <andrew(dot)dunstan(at)2ndquadrant(dot)com>
To: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: PostgreSQL Hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: VACUUM FULL vs dropped columns
Date: 2018-03-10 23:24:28
Message-ID: CAA8=A7_V-_4vudVOqXcWtNcn75VpK0JjtFwsq=VPBLP366Hajg@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Sun, Mar 11, 2018 at 9:49 AM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> Andrew Dunstan <andrew(dot)dunstan(at)2ndquadrant(dot)com> writes:
>> Why does VACUUM FULL cause the size of this table with a single
>> dropped column (1 out of 1000) cause the table size to double?
>
> VACUUM FULL will rewrite the tuples with a null bitmap where they
> had none before (cf reform_and_rewrite_tuple). That's only a rather
> marginal increase in the tuple size, but in this particular example,
> it pushes the tuples from just under half a page to just over, so
> that you go from 2 tuples/page to 1.
>

Aha! Thanks for the explanation.

cheers

andrew

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

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Peter Eisentraut 2018-03-10 23:36:48 Re: disable SSL compression?
Previous Message Tom Lane 2018-03-10 23:19:31 Re: VACUUM FULL vs dropped columns