On Fri, Apr 07, 2023 at 04:14:13AM +0200, Alvaro Herrera wrote: > On 2023-Apr-06, Justin Pryzby wrote:
> > +ERROR: relation "c" already exists > > Do you intend to make an error here ? These still look like mistakes in the tests. > Also, I think these table names may be too generic, and conflict with > other parallel tests, now or in the future. > > > +create table d(a int not null, f1 int) inherits(inh_p3, c); > > +ERROR: relation "d" already exists > Sadly, the binary-upgrade mode is a bit of a mess and thus the > pg_upgrade test is failing.
