Re: [Patch] Don't stop a `vacuumdb --all` on a connection failure

On Mon, Aug 31, 2026 at 10:42 PM Pierre Forstmann
<pierre(dot)forstmann(at)gmail(dot)com> wrote:
> The new status of this patch is: Ready for Committer

I have a few questions about the proposed behavior.

vacuumdb already exits with status 0 when VACUUM skips tables with a warning
due to, e.g., insufficient privileges, as long as no other errors occur.
So, could we treat skipped databases the same way and make this the default,
without adding a new --continue option?

*If* we do need an option, --continue sounds broader than its actual scope.
Using the prefix --skip for the option name is better, following the naming of
the existing option --skip-locked?

Also, additional connections for --jobs still use fail_ok=false in
connect_slot(). So, --continue can still abort the entire run
if an additional connection fails, for example, due to a database connection
limit. Is this intentional?

Regards,

--
Fujii Masao