Re: [PATCH] Stop ALTER SYSTEM from making bad assumptions

Hi,

On 2019-08-05 13:34:39 -0400, Stephen Frost wrote:
> * Andres Freund (andres(at)anarazel(dot)de) wrote:
> > On 2019-08-05 12:43:24 -0400, Tom Lane wrote:
> > > On the other hand, people have also opined that they want full error
> > > checking on the inserted values, and that seems out of reach with less
> > > than a full running system (mumble extensions mumble).
> >
> > I think the error checking ought to be about as complete as the one we
> > do during a normal postmaster startup. Afaict that requires loading
> > shared_preload_library extensions, but does *not* require booting up far
> > enough to run GUC checks in a context with database access.
>
> I'm not following this thread of the discussion.

It's about the future, not v12.

> Where are you thinking this error checking would be happening?

A hypothethical post v12 tool that can set config values with as much
checking as feasible. The IMO most realistic tool to do so is postmaster
itself, similar to it's already existing -C. Boot it up until
shared_preload_libraries have been processed, run check hook(s) for the
new value(s), change postgresql.auto.conf, shutdown.

> You're not suggesting that pg_basebackup perform this error checking
> after it modifies the file, are you..?

Not at the moment, at least.

Greetings,

Andres Freund