Carol Walter <walterc@indiana.edu> writes:
> I'm not sure I understand. If there is a lag time between when
> someone exits the database and when the database "knows" that no one
> is still in it, then that shouldn't be the problem here. I tried a
> number of time over the course of 24 hours to rename the database and
> it always reported that the database was in use. No one was logged in
> but me, but the database still reported that the database was in use.
Hmph. Are you sure nothing was connected? One thing that frequently
trips people up here is uncommitted prepared transactions --- look in
pg_prepared_xacts.
regards, tom lane