Re: pg_upgrade: Make testing different transfer modes easier

From: Daniel Gustafsson <daniel(at)yesql(dot)se>
To: "Shinoda, Noriyoshi (PN Japan FSIP)" <noriyoshi(dot)shinoda(at)hpe(dot)com>
Cc: Peter Eisentraut <peter(dot)eisentraut(at)enterprisedb(dot)com>, PostgreSQL Hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: pg_upgrade: Make testing different transfer modes easier
Date: 2022-12-25 22:11:58
Message-ID: 4BB4EA0D-E818-4ECC-8041-4F46F7E63B30@yesql.se
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

> On 19 Dec 2022, at 01:39, Shinoda, Noriyoshi (PN Japan FSIP) <noriyoshi(dot)shinoda(at)hpe(dot)com> wrote:

> With the addition of --copy option, pg_upgrade now has three possible transfer mode options. Currently, an error does not occur even if multiple transfer modes are specified. For example, we can also run "pg_upgrade --link --copy --clone" command. As discussed in Horiguchi-san's previous email, options like "--mode=link|copy|clone" can prevent this problem.
> The attached patch uses the current implementation and performs a minimum check to prevent multiple transfer modes from being specified.

We typically allow multiple invocations of the same parameter with a
last-one-wins strategy, and only error out when competing *different*
parameters are present. A --mode=<string> parameter can still be added as
syntactic sugar, but multiple-choice parameters is not a commonly used pattern
in postgres utilities (pg_dump/restore and pg_basebackup are ones that come to
mind).

--
Daniel Gustafsson https://vmware.com/

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Michael Paquier 2022-12-25 23:02:28 Re: [BUG] pg_upgrade test fails from older versions.
Previous Message Tom Lane 2022-12-25 20:38:49 Re: Error-safe user functions