Re: Improve pg_sync_replication_slots() to wait for primary to advance

Поиск
Список
Период
Сортировка
От Ajin Cherian
Тема Re: Improve pg_sync_replication_slots() to wait for primary to advance
Дата
Msg-id CAFPTHDbNt7QOWQ+wmX9jZ2QrcBbKQH7P3hV=riBzcegHjhBaAQ@mail.gmail.com
обсуждение исходный текст
Ответ на Re: Improve pg_sync_replication_slots() to wait for primary to advance  (shveta malik <shveta.malik@gmail.com>)
Список pgsql-hackers
On Fri, Sep 26, 2025 at 8:14 PM shveta malik <shveta.malik@gmail.com> wrote:
>
>
> It seems it is missed to be changed perhaps due to bringing back
> previous implementation of slotsync_api_reread_config()
>
> >
> > > b)
> > > There is double space in DETAIL msg: "have changed  since"
> > >
> > > Will it be better to shorten the DETAIL as: 'One or more of
> > > primary_conninfo, primary_slot_name, or hot_standby_feedback were
> > > modified.'
> > >
> >
> > Changed.
> >

Fixed.

>
> This too is missed.
>
> > > 6)
> > > postgres=# SELECT pg_sync_replication_slots();
> > > ERROR:  exiting from slot synchronization as promotion is triggered
> > >
> > > Shall we rephrase it similar to the previous message: 'cannot continue
> > > replication slots synchronization as standby promotion is triggered'
> > >
> > Changed.
> >
> > Attaching patch v14 incorporating the above changes.
> >
>

This was rephrased.

> Few trivial comments:
>
> 1)
> slotsync_api_reread_config():
> + * Returns true if conninfo, primary_slot_name or hot_standby_feedback changed.
>
> This comment is no longer valid, we can remove it.
>

Removed.

> 2)
>   /* We are done with sync, so reset sync flag */
>   reset_syncing_flag();
> +
>
> This extra blank line is not needed.

Fixed.

Attaching patch v15 addressing these comments.

regards,
Ajin Cherian
Fujitsu Australia

Вложения

В списке pgsql-hackers по дате отправления: