Date: Fri, 27 Jun 2014 11:58:18 -0400
From: Alvaro Herrera <alvherre@2ndquadrant.com>
To: Heikki Linnakangas <hlinnakangas@vmware.com>
Cc: Pg Bugs <pgsql-bugs@postgresql.org>
Subject: Re: Bogus pg_class.relminmxid value for pg_database
Message-ID: <20140627155818.GN7340@eldon.alvh.no-ip.org>
References: <52401AEA.9000608@vmware.com>
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="huq684BweRXVnRxX"
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <52401AEA.9000608@vmware.com>
User-Agent: Mutt/1.5.21 (2010-09-15)
Precedence: bulk
Sender: pgsql-bugs-owner@postgresql.org


--huq684BweRXVnRxX
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit

Heikki Linnakangas wrote:
> Right after a fresh initdb on a 9.3 or master database:
> 
> postgres=# select oid,relname,relminmxid from pg_class where
> relname='pg_database';
>  oid  |   relname   | relminmxid
> ------+-------------+------------
>  1262 | pg_database | 4244967297
> (1 row)
> 
> That bogus value seems to appear already some time during initdb.
> The relminmxid for all other tables is 1, which is correct.

I finally figured the source of this issue: it's caused by VACUUM FULL
(executed by initdb while copying template1 into template0): we're
missing a protection that relminmxid doesn't go backwards when swapping
heaps.  We already have such a check for xids.

This is mostly innocuous, but I'm about to push the attached fix all the
same, backpatched to 9.3.  (The extra context lines make the patch very
clear.)

-- 
Álvaro Herrera                http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services

--huq684BweRXVnRxX
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment; filename="backwards-mxid.patch"

diff --git a/src/backend/commands/cluster.c b/src/backend/commands/cluster.c
index 6f87170..46b41c1 100644
--- a/src/backend/commands/cluster.c
+++ b/src/backend/commands/cluster.c
@@ -869,14 +869,20 @@ copy_heap_data(Oid OIDNewHeap, Oid OIDOldHeap, Oid OIDOldIndex,
 	/*
 	 * FreezeXid will become the table's new relfrozenxid, and that mustn't go
 	 * backwards, so take the max.
 	 */
 	if (TransactionIdPrecedes(FreezeXid, OldHeap->rd_rel->relfrozenxid))
 		FreezeXid = OldHeap->rd_rel->relfrozenxid;
 
+	/*
+	 * MultiXactCutoff, similarly, shouldn't go backwards either.
+	 */
+	if (MultiXactIdPrecedes(MultiXactCutoff, OldHeap->rd_rel->relminmxid))
+		MultiXactCutoff = OldHeap->rd_rel->relminmxid;
+
 	/* return selected values to caller */
 	*pFreezeXid = FreezeXid;
 	*pCutoffMulti = MultiXactCutoff;
 
 	/* Remember if it's a system catalog */
 	is_system_catalog = IsSystemRelation(OldHeap);
 

--huq684BweRXVnRxX
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


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

--huq684BweRXVnRxX--