Re: sql_drop Event Triggerg

Robert Haas escribió:
> On Fri, Mar 8, 2013 at 9:18 AM, Alvaro Herrera <alvherre(at)2ndquadrant(dot)com> wrote:
> > Robert Haas escribió:
> >> On Tue, Mar 5, 2013 at 12:45 PM, Alvaro Herrera
> >> <alvherre(at)2ndquadrant(dot)com> wrote:
> >> > Hmm, maybe I should be considering a pair of macros instead --
> >> > UTILITY_START_DROP and UTILITY_END_DROP. I'll give this a try. Other
> >> > ideas are welcome.
> >>
> >> That seems like a possibly promising idea. I do wonder how well any
> >> of this is going to scale.
> >
> > I did followup with a patch implementing that; did you see it?
>
> No, sorry. Which thread is it on?

http://www.postgresql.org/message-id/20130305214218.GP9507@alvh.no-ip.org

I think Gmail's feature of breaking threads when subject changes is an
annoyance here. I somehow added a "g" at the end and later dropped it.
I didn't remember that behavior of Gmail's.

> > The current idea of having a
> > function that returns objects affected by the command seems relatively
> > sensible. For drops, it seems pretty straighforward so far. For CREATE
> > it's probably somewhat more involved, but seems doable in principle (but
> > yes, we're going to have to sprinkle ProcessUtility() with a lot of
> > UTILITY_START/END_CREATE calls).
> >
> > Not sure about ALTER; maybe we will need a completely different idea to
> > attack that.
>
> I am inclined to think that putting this logic in ProcessUtility isn't
> scalable, even for CREATE, and even moreso for ALTER, unless we can
> put it around everything in that function, rather than each command
> individually. Suppose for example that on entry to that function we
> simply did this:
>
> if (isCompleteQuery)
> ++CompleteQueryNestingLevel;
>
> ...and at exit, we did the reverse. This could work a bit like the
> GUC nesting level.

Hmm, this seems an interesting idea to explore.

--
Álvaro Herrera http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services