Horton

Re: logical replication: restart_lsn can go backwards (and more), seems broken since 9.4

Tomas Vondra <tomas@vondra.me>

View thread View on postgr.es
From: Tomas Vondra <tomas@vondra.me>
To: Masahiko Sawada <sawada.mshk@gmail.com>
Cc: PostgreSQL Hackers <pgsql-hackers@postgresql.org>
Date: 2024-11-11T22:10:32Z
Lists: pgsql-hackers 
On 11/11/24 21:56, Masahiko Sawada wrote:
> ...
>>
>>> My hypothesis is that we wanted to prevent such
>>> case by the first if block:
>>>
>>>     /* don't overwrite if have a newer restart lsn */
>>>     if (restart_lsn <= slot->data.restart_lsn)
>>>     {
>>>     }
>>>
>>
>> Yeah, that condition / comment seems to say exactly that.
>>
>> Do you plan / expect to work on fixing this? It seems you proposed the
>> right fix in that old thread, but it's been inactive since 2023/02 :-(
> 
> I'm happy to work on this fix. At that time, I was unsure if my fix
> was really correct and there was no further discussion.
> 

Thanks. I'm not sure about the correctness either, but I think it's
clear the issue is real, and it's not difficult to reproduce it.

regards

-- 
Tomas Vondra

Commits

  1. Fix a possibility of logical replication slot's restart_lsn going backwards.

  2. Introduce logical decoding.