Date: Wed, 24 Jun 2020 11:15:14 -0400
From: Alvaro Herrera <alvherre@2ndquadrant.com>
To: Fujii Masao <masao.fujii@oss.nttdata.com>
Cc: Kyotaro Horiguchi <horikyota.ntt@gmail.com>, amit.kapila16@gmail.com,
	michael@paquier.xyz, pgsql-hackers@postgresql.org
Subject: Re: min_safe_lsn column in pg_replication_slots view
Message-ID: <20200624151514.GA17861@alvherre.pgsql>
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <727f0d64-b720-8f81-0473-56c0278b9a10@oss.nttdata.com>
User-Agent: Mutt/1.10.1 (2018-07-13)
Precedence: bulk

On 2020-Jun-24, Fujii Masao wrote:

> On 2020/06/24 8:39, Alvaro Herrera wrote:

> > I think we should publish the value from wal_keep_segments separately
> > from max_slot_wal_keep_size.  ISTM that the user might decide to change
> > or remove wal_keep_segments and be suddenly at risk of losing slots
> > because of overlooking that it was wal_keep_segments, not
> > max_slot_wal_keep_size, that was protecting them.
> 
> You mean to have two functions that returns
> 
> 1. "current WAL LSN - wal_keep_segments * 16MB"
> 2. "current WAL LSN - max_slot_wal_keep_size"

Hmm, but all the values there are easily findable.  What would be the
point in repeating it?

Maybe we should disregard this line of thinking and go back to
Horiguchi-san's original proposal, to wit use the "distance to
breakage", as also supported now by Amit Kapila[1] (unless I
misunderstand him).

[1] https://postgr.es/m/CAA4eK1L2oJ7T1cESdc5w4J9L3Q_hhvWqTigdAXKfnsJy4=v13w@mail.gmail.com

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