Message-ID: <484F79B0.5060401@archonet.com>
Date: Wed, 11 Jun 2008 08:07:28 +0100
From: Richard Huxton <dev@archonet.com>
User-Agent: Thunderbird 2.0.0.14 (X11/20080421)
MIME-Version: 1.0
To: Richard Broersma <richard.broersma@gmail.com>
CC: PostgreSQL - General ML <pgsql-general@postgresql.org>,
	pgsql-odbc@postgresql.org
Subject: Re: what gives: SELECT INVALID SELECT STATEMENT TO FORCE
	ODBC DRIVER TO UNPREPARED STATE
References: <396486430806102106i4915845ag746566c333bdc684@mail.gmail.com>
In-Reply-To: <396486430806102106i4915845ag746566c333bdc684@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Richard Broersma wrote:
> Server logs below.  Notice how an error raised in a connection to the
> proj02u20411 database forces the transaction in the instrumentation
> database to rollback.
> Can anyone explain why connections that as far as PG should conserned
> are unrelated are actually interferring with one another?

I don't see any errors on proj...

> --------------------------------------------------------------------------------
[snip]
> 2008-06-10 15:04:52instrumentation ERROR:  syntax error at or near
> "SELECT" at character 16
> 2008-06-10 15:04:52instrumentation STATEMENT:  SELECT INVALID SELECT
> STATEMENT TO FORCE ODBC DRIVER TO UNPREPARED STATE
> 2008-06-10 15:04:52instrumentation LOG:  statement: ROLLBACK

That appears to be a deliberate bad query issued by something (above the 
odbc layer?). I think you've got something sending "SELECT INVALID 
SELECT STATEMENT TO FORCE ODBC DRIVER TO UNPREPARED STATE" as a query to 
force a rollback. Why it's not just sending "ROLLBACK" I don't know.

-- 
   Richard Huxton
   Archonet Ltd