Re: Issue with logical replication slot during switchover

On Wed, Oct 8, 2025 at 12:27 PM shveta malik <shveta(dot)malik(at)gmail(dot)com> wrote:

> On Fri, Oct 3, 2025 at 1:28 PM Fabrice Chapuis <fabrice636861(at)gmail(dot)com>
> wrote:
> >
> > Hi Shveta,
> > Here is the v4 of the patch with pg_replication_slots view modified to
> display the field allow_overwrite. Doc was also updated.
> >
>
> The patch looks okay. The parameter name is still open for discussion,
> and the comments could be improved. But we can focus on these finer
> details once more reviewers start reviewing and there’s general
> agreement on the concept.
>
> One trivial comment: we can slightly modify the doc to have something like
> this:
>
> This parameter controls whether an existing logical replication slot
> on the standby (with synced=false) can be overwritten during logical
> replication slot synchronization (see Section 47.2.3). The default is
> false. When true, an existing user slot with the same name on the
> standby will be synchronized using the primary’s failover slot.
>
> <please see high-availability.sgml to find how 'Section 47.2.3' can be
> referenced in the doc>
> ~~
>
> The next step will be to provide a way to modify this parameter via an
> alter API, say pg_alter_logical_replication_slot(). This API can later
> be extended to handle other parameters. This API can be implemented in
> patch002 for easier review.
>
> thanks
> Shveta
>

Hi,
Here is the patch V5, I change with your doc text proposition and the
link.
At this stage, the patch can be submitted to the current commit fest for
review?

With Regards,

Fabrice