Re: Resetting recovery target parameters in pg_createsubscriber

Alena Vinter <dlaaren8@gmail.com>

From: Alena Vinter <dlaaren8@gmail.com>
To: Alexander Korotkov <aekorotkov@gmail.com>
Cc: Robert Haas <robertmhaas@gmail.com>, Michael Paquier <michael@paquier.xyz>, Ilyasov Ian <ianilyasov@outlook.com>, "Hayato Kuroda (Fujitsu)" <kuroda.hayato@fujitsu.com>, PostgreSQL Hackers <pgsql-hackers@postgresql.org>
Date: 2025-11-10T05:02:14Z
Lists: pgsql-hackers
On Tue, 4 Nov 2025 at 04:28, Alexander Korotkov <aekorotkov@gmail.com>
wrote:

> I have rechecked this.  It appears that pg_createsubscriber writes the
> recovery configuration to the output and only in verbose mode.  So,
> it's far no guaranteed that this information would be accessible.  One
> may run pg_createsubscriber not in verbose mode or don't save its
> output.  I suggest we should re-implement this in a way Michael
> proposed [1]: save the configuration to pg_createsubscriber.conf.old
> file.
>

Alexander, I'm not in favor of saving additional files. This approach seems
to replace one type of leftover artifact (recovery params) with another
(debug-files). Neither option is good.
As Michael pointed out, the parameters might be useful for post-debugging
purposes. This suggests to me that they are, by nature, debugging
information. Therefore, it seems appropriate that they should be captured
by the verbose/debug mode. If verbose mode isn't used, we lose more than
just the recovery parameters — we also lose the sequence of commands for
managing replication slots and other steps. Following this logic, why not
save all information in non-verbose mode that might be used for debugging?


Robert, I'll think more about a valid scenario (including the one you
proposed) and get back with results later.

---
Regards,
Alena Vinter