Tom Lane writes:
> What I'm thinking is that if we hard-wired usesysid = 1 for the
> superuser, it'd be possible to arrange for standalone backends to fire
> up with that sysid and superuserness assumed, and not consult pg_shadow
> at all. Then you'd have a platform in which you could do CREATE USER.
I had always figured that you could use bki to recover from these things,
but a quick attempt shows that you can't.
You proposal makes sense from a Unix admin point of view (booting into
single user mode without password). Since we have a check against root
access and against too liberal PGDATA permissions, I think this would be
safe. Possibly we need to guard against setgid problems as well.
> Next mind-bending problem: recover from DROP TABLE pg_class ;-)
Definitely BKI land. But that usecatupd field does make some sense,
apparently.
--
Peter Eisentraut peter_e@gmx.net http://funkturm.homeip.net/~peter