I've been unable to time a difference here on a DB with 20 tables -- but
perhaps you will be able to on whatever box you ran the first one on. I
didn't notice the initial version was slow -- but can certainly see how
it would be on a large db.
Scan of pg_depend for attributes using the domain. Each resulting
attribute is double checked.
I'll work out the locking aspect shortly.
Rod Taylor <rbt(at)rbt(dot)ca>
PGP Key: http://www.rbt.ca/rbtpub.asc
pgsql-patches by date
|Next:||From: Peter Eisentraut||Date: 2002-12-13 19:55:16|
|Subject: Re: [BUGS] Bug #718: request for improvement of /? to show|
|Previous:||From: Bruce Momjian||Date: 2002-12-13 17:02:06|
|Subject: Re: postmaster segfaults when pg_hba.cof is missing|