Re: Problems with pg_upgrade and extensions referencing catalog tables/views

"Nasby, Jim" <nasbyj(at)amazon(dot)com> writes:
> The problem is that pg_dump --binary-upgrade intentionally does not
> simply issue a `CREATE EXTENSION` command the way a normal dump does, so
> that it can control the OIDs that are assigned to objects[1].

That's not the only reason. The original concerns were about not
breaking the extension, in case the destination server had a different
version of the extension available. CREATE EXTENSION doesn't normally
guarantee that you get an exactly compatible extension version, which
is a good thing for regular pg_dump and restore but a bad thing
for binary upgrade.

I'm not really sure how to improve the situation you describe, but
"issue CREATE EXTENSION and pray" doesn't sound like a solution.

regards, tom lane