Re: [PATCH] Explain what the default output_plugin_libraries do - Mailing list pgsql-hackers

From Jacob Champion
Subject Re: [PATCH] Explain what the default output_plugin_libraries do
Date
Msg-id CAOYmi+nQ=+Ss8G0aOTVc7Ss41iqhsN3_OqnTUYCWKyS9ahyXPw@mail.gmail.com
Whole thread
In response to RE: [PATCH] Explain what the default output_plugin_libraries do  ("Hayato Kuroda (Fujitsu)" <kuroda.hayato@fujitsu.com>)
Responses Re: [PATCH] Explain what the default output_plugin_libraries do
RE: [PATCH] Explain what the default output_plugin_libraries do
RE: [PATCH] Explain what the default output_plugin_libraries do
List pgsql-hackers
On Thu, Sep 17, 2026 at 1:42 AM Hayato Kuroda (Fujitsu)
<kuroda.hayato@fujitsu.com> wrote:
> How about attached?

> +    An output plugin must be installed as a dynamically loadable library in a
> +    location where the server can find it (see <xref linkend="xfunc-c-dynload"/>).

I think we should explicitly recommend installation into the
dynamic_library_path ($libdir for most people). Anything outside of
dynamic_library_path couples clients to the directory layout, and it
makes output_plugin_libraries maintenance harder/uglier.

> > Okay. Any thoughts on Bharath's note on "publication slot creation" above?
>
> I could not find the term "publication slot" in the existing doc, so I prefer
> his idea. Another one I came up with to just simplify:
>
> ```
> Removing from output_plugin_libraries will cause future logical replication to
> fail with an error. (Beware that existing replication tools may not degrade gracefully
> if the server is configured in this way.)

But only built-in logical replication will fail. How about

             Core plugin for logical replication. Removing
<literal>pgoutput</literal>
             from <literal>output_plugin_libraries</literal> will cause future
             subscription connections, and maintenance of the replication slots
             for subscriptions, to fail with an error. (Beware that ...

?

Thanks!
--Jacob



pgsql-hackers by date:

Previous
From: Alexandre Felipe
Date:
Subject: Re: Restructured Shared Buffer Hash Table
Next
From: Melanie Plageman
Date:
Subject: Re: eliminate xl_heap_visible to reduce WAL (and eventually set VM on-access)