Re: BUG #17641: postgresql service persistently restarts because "could not open shared memory segment"

From: Andres Freund <andres(at)anarazel(dot)de>
To: pgsql-bugs(at)lists(dot)postgresql(dot)org, Thomas Munro <thomas(dot)munro(at)gmail(dot)com>, j(dot)falch(dot)16(at)gmail(dot)com
Cc: Sandeep Thakkar <sandeep(dot)thakkar(at)enterprisedb(dot)com>
Subject: Re: BUG #17641: postgresql service persistently restarts because "could not open shared memory segment"
Date: 2022-10-16 21:11:55
Message-ID: B24B6EC7-65E9-4D01-B888-6A2EA3960DD1@anarazel.de
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

Hi,

On October 16, 2022 2:05:59 PM PDT, Thomas Munro <thomas(dot)munro(at)gmail(dot)com> wrote:
>On Sun, Oct 16, 2022 at 4:31 PM Thomas Munro <thomas(dot)munro(at)gmail(dot)com> wrote:
>> On Sun, Oct 16, 2022 at 8:52 AM PG Bug reporting form
>> <noreply(at)postgresql(dot)org> wrote:
>> > 2022-10-15 20:00:37.973 CEST [19124] LOG: Datenbanksystem ist bereit, um
>> > Verbindungen anzunehmen
>> > 2022-10-15 20:00:44.113 CEST [23816] FATAL: konnte Shared-Memory-Segment
>> > »Global/PostgreSQL.1886284784« nicht öffnen: No such file or directory
>> > 2022-10-15 20:00:44.113 CEST [23816] FATAL: can't attach the same segment
>> > more than once
>>
>> Hi,
>>
>> Do you have any extensions installed, and can you share any
>> non-default settings from postgresql.conf or similar?
>
>Off-list reply was that this happened straight away with the
>EnterpriseDB installer without changing any settings, using
>PGDATA=C:\ProgramData\PostgreSQL\data. I can't immediately guess how
>that first error is caused, but it's also interesting that we
>apparently screwed up our internal state afterwards because the second
>implies that we tried to attach to something we already knew we were
>attached to. But we haven't had any reports like this from our
>Windows-based developers who've been testing the 15 branch for a
>while, and we don't see this on the buildfarm or CI, so I think the
>first port of call here will be to ping the EDB package team and see
>if they can reproduce or understand this? CCing Sandeep for info.

I wonder if there's antivirus active, and if so, which one?

I suspect the problem with the two FATALs might be due to falling in the middle of attaching to a segment, and then trying to attach again during the proc exit processing.

Andres
--
Sent from my Android device with K-9 Mail. Please excuse my brevity.

In response to

Browse pgsql-bugs by date

  From Date Subject
Next Message Tom Lane 2022-10-16 23:21:50 Re: BUG #17644: EXPLAIN VERBOSE fails on query with SEARCH BREADTH FIRST
Previous Message Thomas Munro 2022-10-16 21:05:59 Re: BUG #17641: postgresql service persistently restarts because "could not open shared memory segment"