Horton

Re: [BUG] Logical replica crash if there was an error in a function.

Tom Lane <tgl@sss.pgh.pa.us>

View thread View on postgr.es
From: Tom Lane <tgl@sss.pgh.pa.us>
To: "Anton A. Melnikov" <aamelnikov@inbox.ru>
Cc: vignesh C <vignesh21@gmail.com>, Andres Freund <andres@anarazel.de>, Amit Kapila <amit.kapila16@gmail.com>, Masahiko Sawada <sawada.mshk@gmail.com>, Michael Paquier <michael@paquier.xyz>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>, Maxim Orlov <orlovmg@gmail.com>
Date: 2023-04-05T14:35:59Z
Lists: pgsql-hackers 
"Anton A. Melnikov" <aamelnikov@inbox.ru> writes:
> On 03.04.2023 21:49, Tom Lane wrote:
>> I did not think this case was worth memorializing in a test before,
>> and I still do not.  I'm inclined to reject this patch.

> Could you help me to figure out, please.

The problem was an Assert that was speculative when it went in,
and which we eventually found was wrong in the context of logical
replication.  We removed the Assert.  I don't think we need a test
case to keep us from putting back the Assert.  That line of thinking
leads to test suites that run for fourteen hours and are near useless
because developers can't run them easily.

			regards, tom lane

Commits

  1. Avoid crash after function syntax error in a replication worker.