Re: pg_upgrade, tables_with_oids.txt -> tables_with_oids.sql? - Mailing list pgsql-hackers

From Daniel Gustafsson
Subject Re: pg_upgrade, tables_with_oids.txt -> tables_with_oids.sql?
Date
Msg-id 825FC57F-2B2D-4E54-BB21-EDCA91FADCC5@yesql.se
Whole thread Raw
In response to pg_upgrade, tables_with_oids.txt -> tables_with_oids.sql?  ("Daniel Westermann (DWE)" <daniel.westermann@dbi-services.com>)
List pgsql-hackers
> On 6 Nov 2022, at 09:48, Daniel Westermann (DWE) <daniel.westermann@dbi-services.com> wrote:

> as I've just upgraded an instance which contained tables "WITH OIDS" I wonder if it would make sense if pg_upgrade
directlycreates a script to fix those. I know it is easy to that with e.g. sed over tables_with_oids.txt but it would
bemore convenient to have the script generated directly. 

For the checks on the old system we don't generate any scripts, only reports of
problems.  I don't recall the reasoning but I would assume it stems from some
checks being up to the user to deal with, no one-size-fits-all script is
possible.  Having them all generate reports rather than scripts makes that
consistent across the old checks.

In this particular case we probably could safely make a script, but if we we'd
need to expand testing to validate it etc so I'm not sure it's worth it.

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




pgsql-hackers by date:

Previous
From: Ted Yu
Date:
Subject: Re: [CAUTION!! freemail] Re: Partial aggregates pushdown
Next
From: Maxim Orlov
Date:
Subject: Re: [PATCH] Add initial xid/mxid/mxoff to initdb