Re: BUG #17717: Regression in vacuumdb (15 is slower than 10/11 and possible memory issue)

I wrote:
> Justin Pryzby <pryzby@telsasoft.com> writes:
>> I assumed it would look like:
>> VACUUM (UPDATE_DATABASE_STATS {yes,no,only})

> Meh.  We could do it like that, but I think options that look like
> booleans but aren't are messy.

Note that I'm not necessarily objecting to there being just one option,
only to its values being a superset-of-boolean.  Perhaps this'd work:

VACUUM (DATABASE_STATS {UPDATE,SKIP,ONLY})

            regards, tom lane