Re: BUG #6497: Error sent to client, but data written anyway

From: Tatsuo Ishii <ishii(at)postgresql(dot)org>
To: tgl(at)sss(dot)pgh(dot)pa(dot)us
Cc: rlowe(at)pablowe(dot)net, Kevin(dot)Grittner(at)wicourts(dot)gov, pgsql-bugs(at)postgresql(dot)org
Subject: Re: BUG #6497: Error sent to client, but data written anyway
Date: 2012-02-29 23:10:02
Message-ID: 20120301.081002.242108456694127118.t-ishii@sraoss.co.jp
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

> You might wish to believe that you can ignore the problem, but you can't.
> No matter what Postgres does or doesn't do, external issues such as
> network failures can create the problem of a transaction possibly being
> committed while the client remains in doubt whether it happened or not.

I think this is a well known issue and the only solution is a
transaction system covering the communication path. Since there's no
such a transaction system exists in PostgreSQL, clients should be very
carefull when try to resend DMLs as you already pointed out.

F.Y.I. pgpool-II disconnects the connection to PostgreSQL upon failure
including network problem, rather than trying to resend packet to
PostgreSQL due to the reason stated above.
--
Tatsuo Ishii
SRA OSS, Inc. Japan
English: http://www.sraoss.co.jp/index_en.php
Japanese: http://www.sraoss.co.jp

In response to

Browse pgsql-bugs by date

  From Date Subject
Next Message sujatha.pelluru 2012-03-01 10:58:08 BUG #6499: query to get data with single quote
Previous Message k/a 2012-02-29 22:25:30 Re: BUG #6498: with recursive / union all