Date: Tue, 27 Nov 2018 12:38:06 -0800
From: Andres Freund <andres@anarazel.de>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Andrew Dunstan <andrew.dunstan@2ndquadrant.com>,
	pgsql-committers@lists.postgresql.org
Subject: Re: pgsql: Update pg_upgrade test for reg* to include regrole and
 regnamesp
Message-ID: <20181127203806.behekoj55rnwuuvg@alap3.anarazel.de>
References: <E1gRS1V-0007eS-1p@gemulon.postgresql.org>
 <8863.1543297423@sss.pgh.pa.us>
 <20181127063226.mpyijg3jwrmmm3rr@alap3.anarazel.de>
 <6f5edd28-e55f-7399-d89c-d2d1ed2d1e51@2ndQuadrant.com>
 <2735.1543333649@sss.pgh.pa.us>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <2735.1543333649@sss.pgh.pa.us>
Precedence: bulk

Hi,

On 2018-11-27 10:47:29 -0500, Tom Lane wrote:
> Andrew Dunstan <andrew.dunstan@2ndquadrant.com> writes:
> > In any case, I think we will probably need to bite the bullet and have 
> > pg_dump render LOs in OID order.
> 
> +1.  Looking at the relevant code:
> 
>     /*
>      * Currently, we re-fetch all BLOB OIDs using a cursor.  Consider scanning
>      * the already-in-memory dumpable objects instead...
>      */
>     if (fout->remoteVersion >= 90000)
>         blobQry = "DECLARE bloboid CURSOR FOR SELECT oid FROM pg_largeobject_metadata";
>     else
>         blobQry = "DECLARE bloboid CURSOR FOR SELECT DISTINCT loid FROM pg_largeobject";

Done. Thanks for chiming in.

Andrew, the <= 9.2 failure https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=crake&dt=2018-11-27%2018%3A18%3A21
seems to be related to nabstime etc not being available anymore.  We've
not backpatched

Author: Tom Lane <tgl@sss.pgh.pa.us>
Branch: REL_11_STABLE Release: REL_11_0 [184951a48] 2018-10-12 19:33:56 -0400
Branch: REL_10_STABLE Release: REL_10_6 [9320263ae] 2018-10-12 19:33:56 -0400
Branch: REL9_6_STABLE Release: REL9_6_11 [2ad422ce1] 2018-10-12 19:33:56 -0400
Branch: REL9_5_STABLE Release: REL9_5_15 [43cc4e49e] 2018-10-12 19:33:56 -0400
Branch: REL9_4_STABLE Release: REL9_4_20 [7b88c1ddd] 2018-10-12 19:33:57 -0400
Branch: REL9_3_STABLE Release: REL9_3_25 [fb583c80d] 2018-10-12 19:33:57 -0400

    Remove abstime, reltime, tinterval tables from old regression databases.

that far.  I've locally verified that upgrade works after

SELECT format('ALTER TABLE %s DROP COLUMN %s;', attrelid::regclass, attname) FROM pg_attribute WHERE attrelid > 16384 AND atttypid IN ('abstime'::regtype, 'reltime'::regtype, 'tinterval'::regtype)
\gexec

or something in that vein.

Greetings,

Andres Freund