Bruce Momjian <pgman@candle.pha.pa.us> writes:
>> ... On top of that, that's also the risk of someone being a
>> superuser. They will ALWAYS have the power to hose things. Period. As
>> such, I don't consider that to be a valid argument.
> That was my feeling too. If you can't trust the other admins, it is
> hard for us to trust them either.
Sigh. It's not about trust: it's about whether pg_upgrade can enforce
or at least check its assumptions. I don't feel that it's a
production-grade tool as long as it has to cross its fingers that the
DBA made no mistakes.
Also, if the previous example had no impact on you, try this one:
$ postmaster -N 1 -c superuser_reserved_connections=0 &
$ pg_dumpall
pg_dump: [archiver (db)] connection to database "regression" failed: FATAL: Sorry, too many clients already
pg_dumpall: pg_dump failed on regression, exiting
$
-N 1 *will* cause problems.
regards, tom lane