Hi Jean-Christophe,
On Mon, 28 Jul 2025 10:23:07 +0200
Jean-Christophe Arnu <jca...@gmail.com> wrote:
> Le lun. 28 juil. 2025 à 10:01, Jean-Christophe Arnu <jca...@gmail.com> a
> écrit :
>
> > Here's a first version of this tiny doc patch.
> >
> > Hope this is clear enough.
> >
>
> Sorry, the previous version had a bad wording for one file. Please do not
> review it. The following version is the correct one.
a signal (other than <systemitem>SIGTERM</systemitem>, which is used as
part of a database server shutdown) or an error by the shell (such as
command not found), then recovery will abort and the server will not
start
- up.
+ up. However, the server will also not start if the command returns a code
+ of 128 and above.
It seems redundant with the explanation in this paragraph when you know that a
code greater than 125 is returned on shell error or signal.
As I'm sure you already know, this behavior is documented on the
archive_command side using these words:
«
When the archive command is terminated by a signal (other than SIGTERM that
is used as part of a server shutdown) or an error by the shell with an exit
status greater than 125 (such as command not found), or if the archive
function emits an ERROR or FATAL, the archiver process aborts and gets
restarted by the postmaster.
»
So I assume we could keep the same documentation style for the restore_command
side:
«
An exception is that if the command was terminated by a signal (other than
SIGTERM, which is used as part of a database server shutdown) or an error by
the shell **with an exit status greater than 125** (such as command not
found), then recovery will abort and the server will not start up.
»
What do you think?
