They're all:

NOT DEFERRABLE INITIALLY IMMEDIATE FOR EACH ROW EXECUTE PROCEDURE ...

according to pg_constraint/pg_trigger none of the constraints are
deferrable.

Matt

On Tue, 2005-05-10 at 07:59 -0400, Jan Wieck wrote:
> On 5/9/2005 11:53 PM, Matthew Rich wrote:
> > Hi,
> > 	I keep seeing this strange behavior on a slave when running scripts
> > through "execute script" that include a delete statement (on a table
> > that has PK/FK constraints).
> > 
> > PGRES_FATAL_ERROR ERROR:  could not find trigger 57715371
> > 
> > this appears to be an error generated in the postgres backend
> > DeferredTriggerExecute().  I could speculate that an attempt is being
> > made at executing the trigger after slony has already disabled it again.
> 
> If there is by chance any DEFERRED constraint, then I would agree with 
> your speculation. DEFERRED constraints are executed at COMMIT time and 
> the "execute script" in fact has removed the constraint triggers again 
> by then.
> 
> 
> Jan
> 
> > 
> > I suspect that it is actually not advisable to do DML through the
> > execute script command, though the documentation doesn't say anything
> > about it.  The script that generates this error runs normally on the
> > master though I have seen this same error generated from a different
> > script on the master.
> > 
> > Has anyone had this problem?  I couldn't find any information on
> > anything even close, though perhaps most users don't try to run DML.
> > My problem is that I need to run the script as a single transaction and
> > it includes alterations to the schema and the data, so I can't separate
> > the changes.  Any help is appreciated. :)
> > 
> > Thanks,
> > 
> > Matt
> > 
> > _______________________________________________
> > Slony1-general mailing list
> > Slony1-general at gborg.postgresql.org
> > http://gborg.postgresql.org/mailman/listinfo/slony1-general
> 
> 
-- 
Matthew Rich <mrich at tigris.org>