Message-ID: <54086B77.9090102@aklaver.com>
Date: Thu, 04 Sep 2014 06:39:03 -0700
From: Adrian Klaver <adrian.klaver@aklaver.com>
User-Agent: Mozilla/5.0 (X11; Linux i686;
 rv:24.0) Gecko/20100101 Thunderbird/24.6.0
MIME-Version: 1.0
To: Alanoly Andrews <alanolya@invera.com>,
 'Craig James' <cjames@emolecules.com>
CC: "pgsql-admin@postgresql.org" <pgsql-admin@postgresql.org>,
 "pgsql-general@postgresql.org" <pgsql-general@postgresql.org>
Subject: Re: [GENERAL] Re: Cannot retrieve images inserted through
 VB and odbc, after a table reorganization.
References: <6909b219ecae4d41b431efdaf6c953a3@Exch2.invera.com>
	<CAFwQ8rd0Abfm=0j4Oq5D0hcRPBn=D7A-HJzSPJmDk=EgDMA9zQ@mail.gmail.com>
	<a45db4656040438eaf2034a8f5870b8d@Exch2.invera.com>
	<5407679B.2030906@aklaver.com>
 <aafcb3551e2c4dddb696ec4a6d6087e8@Exch2.invera.com>
In-Reply-To: <aafcb3551e2c4dddb696ec4a6d6087e8@Exch2.invera.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk
Sender: pgsql-admin-owner@postgresql.org

On 09/04/2014 05:56 AM, Alanoly Andrews wrote:
> Hi Adrian,
>
> Thanks for that explanation of how the "relfilenode" changes after a table reorganization. It is not surprising that this happens because the table rows are being physically moved from one location to another. But such changes at the backend should be transparent to the end user. The VB code at the client side runs a simple sql like "select image from image_table where image_key=somevalue". There is no reference to postgres-specific internal variables like "oid" and "relfilenode". I do not know the inner workings of the postgres odbc driver; but I would be surprised if it works at the granularity of "oid" and "relfilenode" and that it would store the actual physical values of relfilenode (which would keep changing after every table reload, reorg etc.).

In addition to what David mentioned, some more detail on what you are 
calling a 'large object'.  In Postgres there is not really a 'large 
object' type(though I am guilty of saying there is), instead there is a 
way of working with large objects outside of the bytea type. Now in the 
ODBC FAQ there is a recipe for creating a 'lo' type:

http://psqlodbc.projects.pgfoundry.org/faq.html#4.4

Is this what you are using?

If not how are you working with the large objects?

Also given that you are working with the Postgres ODBC driver, you might 
want to break the cross post rule and ask this question on the psql-odbc 
list:

http://www.postgresql.org/list/pgsql-odbc/

In any case your ODBC settings would be helpful.

>
> Alanoly.
>


-- 
Adrian Klaver
adrian.klaver@aklaver.com


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