RE: logical_replication_mode

On Friday, August 25, 2023 5:56 PM Amit Kapila <amit(dot)kapila16(at)gmail(dot)com> wrote:
>
> On Fri, Aug 25, 2023 at 12:38 PM Peter Eisentraut <peter(at)eisentraut(dot)org> wrote:
> >
> > On 25.08.23 08:52, Zhijie Hou (Fujitsu) wrote:
> > > On Friday, August 25, 2023 12:28 PM Amit Kapila
> <amit(dot)kapila16(at)gmail(dot)com> wrote:
> > >>
> > >> On Thu, Aug 24, 2023 at 12:45 PM Peter Eisentraut
> > >> <peter(at)eisentraut(dot)org>
> > >> wrote:
> > >>>
> > >>> I suggest we rename this setting to something starting with debug_.
> > >>> Right now, the name looks much too tempting for users to fiddle with.
> > >>> I think this is similar to force_parallel_mode.
> > >>>
> > >>
> > >> +1. How about debug_logical_replication?
> > >>
> > >>> Also, the descriptions in guc_tables.c could be improved. For
> > >>> example,
> > >>>
> > >>> gettext_noop("Controls when to replicate or apply each
> > >>> change."),
> > >>>
> > >>> is pretty content-free and unhelpful.
> > >>>
> > >>
> > >> The other possibility I could think of is to change short_desc as:
> > >> "Allows to replicate each change for large transactions.". Do you
> > >> have any better ideas?
> > >
> > > How about "Forces immediate streaming or serialization of changes in
> > > large transactions." which is similar to the description in document.
> > >
> > > I agree that renaming it to debug_xx would be better and here is a
> > > patch that tries to do this.
> >
> > Maybe debug_logical_replication is too general? Something like
> > debug_logical_replication_streaming would be more concrete.
> >
>
> Yeah, that sounds better.

OK, here is the debug_logical_replication_streaming version.

Best Regards,
Hou zj