"David G. Johnston" <david.g.johnston@xxxxxxxxx> writes: > As an aside the alter user/database commands do end up requiring the user > to disconnect and reconnect. Is there a hard limitation why an > administrator can't send some kind of signal to cause a re-read of those by > an active session? If we wanted to redefine their behavior, we could probably make that happen, though it's not clear we want to. For example, if you've done SET ROLE, should a reload of ALTER USER values pertaining to your original role affect you? Another interesting point is that the config file re-read behavior will currently happen at next command receipt, even if you are inside an aborted transaction. We cannot read the system catalogs if the current transaction is aborted, so the timing would have to be subtly different in any case. regards, tom lane -- Sent via pgsql-general mailing list (pgsql-general@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general
