From: Tom Lane <tgl@sss.pgh.pa.us>
To: Bruce Momjian <bruce@momjian.us>
cc: Rural Hunter <ruralhunter@gmail.com>,
	PostgreSQL-development <pgsql-hackers@postgresql.org>
Subject: Re: [ADMIN] pg_upgrade from 9.1.3 to 9.2 failed
In-reply-to: <20120917043500.GB20123@momjian.us>
References: <5052C402.8020003@gmail.com> <20120914142616.GA11265@momjian.us>
	<50535302.8080000@gmail.com> <20120914183915.GA5933@momjian.us>
	<5053F896.60005@gmail.com> <20120915180603.GA20907@momjian.us>
	<505557CD.8090304@gmail.com> <20120916171746.GA24907@momjian.us>
	<50567E60.2000005@gmail.com> <2932.1347846538@sss.pgh.pa.us>
	<20120917043500.GB20123@momjian.us>
Comments: In-reply-to Bruce Momjian <bruce@momjian.us>
	message dated "Mon, 17 Sep 2012 00:35:00 -0400"
Date: Mon, 17 Sep 2012 00:47:11 -0400
Message-ID: <6578.1347857231@sss.pgh.pa.us>

Bruce Momjian <bruce@momjian.us> writes:
> On Sun, Sep 16, 2012 at 09:48:58PM -0400, Tom Lane wrote:
>> Well, that's even stranger, because (1) information_schema.sql_features
>> ought to have a toast table in either version, and (2) neither pg_dump
>> nor pg_upgrade ought to be attempting to dump or transfer that table.

> I bet pg_upgrade is picking it up from the old cluster because it has an
> oid >= FirstNormalObjectId and the table is not in the information
> schema.

If it *isn't* in information_schema, but is just some random table that
happens to be named sql_features, then it's hard to explain why there's
anything going wrong at all.  My money is on the OP having done a reload
of the information_schema (as per, eg, the release notes for 9.1.2), and
somehow that's confusing pg_dump and/or pg_upgrade.

			regards, tom lane