RE: [PoC] pg_upgrade: allow to upgrade publisher node - Mailing list pgsql-hackers

From Hayato Kuroda (Fujitsu)
Subject RE: [PoC] pg_upgrade: allow to upgrade publisher node
Date
Msg-id TYAPR01MB5866EEF603F76F34AA5D516CF5749@TYAPR01MB5866.jpnprd01.prod.outlook.com
Whole thread Raw
In response to RE: [PoC] pg_upgrade: allow to upgrade publisher node  ("Wei Wang (Fujitsu)" <wangw.fnst@fujitsu.com>)
List pgsql-hackers
Dear Wang,

> I'm not sure if there is any reason to not expose this new option? Do we have
> concerns that users who use this new option by mistake may cause data
> inconsistencies?
>
> BTW, I think that all options of pg_dump (please see the array of long_options
> in the main function of the pg_dump.c file) are currently exposed to the user.

Apart from another database object, --logical-replication-slot-only does not provide
the "perfect" copy. As you might know, some attributes like xmin and restart_lsn
are not copied, it just creates similar replication slots which have same name,
plugin, and options. I think these things may be confused for users.

Moreover, I cannot come up with use-case which DBAs use the option alone.
If there is a good one, I can decide to remove the limitation.

Best Regards,
Hayato Kuroda
FUJITSU LIMITED


pgsql-hackers by date:

Previous
From: Michael Paquier
Date:
Subject: Re: WAL Insertion Lock Improvements
Next
From: Yurii Rashkovskii
Date:
Subject: Re: [PATCH] Allow Postgres to pick an unused port to listen