Re: BUG #17070: Sometimes copy from ingnores transaction

From: Noah Misch <noah(at)leadboat(dot)com>
To: Дмитрий Секретов <d(dot)sekretov(at)gmail(dot)com>
Cc: pgsql-bugs(at)lists(dot)postgresql(dot)org
Subject: Re: BUG #17070: Sometimes copy from ingnores transaction
Date: 2021-09-18 03:18:49
Message-ID: 20210918031849.GA2911957@rfd.leadboat.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

On Fri, Sep 17, 2021 at 02:13:43PM +0300, Дмитрий Секретов wrote:
> It happens very rarely and it happens in production (2-3 times a year), i
> can't log all. I look into code and we always start transaction. We have
> tests - rollback works fine. But it happened.

A missing BEGIN is something like one hundred times as likely as a PostgreSQL
bug here. I understand you can't set log_statement=all in production. If you
do a non-production test with log_statement=all, you may be able to locate the
missing BEGIN.

In response to

Browse pgsql-bugs by date

  From Date Subject
Next Message Jaime Casanova 2021-09-18 22:20:17 Re: BUG #17061: Impossible to query the fields of the tuple created by SEARCH BREADTH FIRST BY .. SET ..
Previous Message PG Bug reporting form 2021-09-17 12:17:36 BUG #17194: Issue with pgoutput