Re: [HACKERS] SUBSCRIPTIONS and pg_upgrade - Mailing list pgsql-hackers

From Stephen Frost
Subject Re: [HACKERS] SUBSCRIPTIONS and pg_upgrade
Date
Msg-id 20170417195441.GE9812@tamriel.snowman.net
Whole thread Raw
In response to Re: [HACKERS] SUBSCRIPTIONS and pg_upgrade  (Peter Eisentraut <peter.eisentraut@2ndquadrant.com>)
List pgsql-hackers
Peter,

* Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:
> On 4/13/17 12:11, Robert Haas wrote:
> > I wonder if we should have an --no-subscriptions option, now that they
> > are dumped by default, just like we have --no-blobs, --no-owner,
> > --no-password, --no-privileges, --no-acl, --no-tablespaces, and
> > --no-security-labels.  It seems like there is probably a fairly large
> > use case for excluding subscriptions even if you have sufficient
> > permissions to dump them.
>
> What purpose would that serve in practice?  The subscriptions are now
> dumped disabled, so if they hang around, there is not much impact.

Will they be allowed to be created as a non-superuser when loading into
another cluster?  Even if you might have superuser rights on one
cluster, you may not on another.

> Perhaps an option that also omits publications would make sense?

Yes.

> Or a general filtering facility based on the object tag?

Perhaps..  I'd hate for it to end up being overly complicated though.
The general '--no-security-lables', '--no-tablespaces' and other options
are pretty straight-forward and I think that's a good thing.

Thanks!

Stephen

pgsql-hackers by date:

Previous
From: Stephen Frost
Date:
Subject: Re: [HACKERS] Row Level Security UPDATE Confusion
Next
From: Bruce Momjian
Date:
Subject: Re: [HACKERS] Self-signed certificate instructions