From: Tom Lane <tgl@sss.pgh.pa.us>
To: Andrew Dunstan <andrew.dunstan@2ndquadrant.com>
cc: pgsql-committers@postgresql.org
Subject: Re: pgsql: Support arrays over domains.
In-reply-to: <678e49f8-e32a-9b96-8e3f-a8c63d2f0e18@2ndQuadrant.com>
References: <E1dyLlE-0002gT-H5@gemulon.postgresql.org>
 <678e49f8-e32a-9b96-8e3f-a8c63d2f0e18@2ndQuadrant.com>
Comments: In-reply-to Andrew Dunstan <andrew.dunstan@2ndquadrant.com>
	message dated "Sat, 30 Sep 2017 15:19:05 -0400"
MIME-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-ID: <24122.1506799963.1@sss.pgh.pa.us>
Content-Transfer-Encoding: 8bit
Date: Sat, 30 Sep 2017 15:32:43 -0400
Message-ID: <24123.1506799963@sss.pgh.pa.us>
Precedence: bulk
Sender: pgsql-committers-owner@postgresql.org

Andrew Dunstan <andrew.dunstan@2ndquadrant.com> writes:
> We seem to have caused a problem with pg_upgrade :-(

>     pg_restore: [archiver (db)] could not execute query: ERROR:  pg_type
>     array OID value not set when in binary upgrade mode

> Here we're upgrading from 9.2 to HEAD. Of course, there isn't an array
> oid for the domain in 9.2.

Hmm.  That's pretty annoying.  We could certainly go ahead and assign
some unused type OID to the new array type, but the receiving backend
can't really do that, since it doesn't know which type OIDs will be
assigned later in the dump file.

Maybe we could put the burden on pg_upgrade (pg_dump --binary-upgrade)
to select an OID that's not in use in the source database.

			regards, tom lane


-- 
Sent via pgsql-committers mailing list (pgsql-committers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-committers