Vick Khera vivek at khera.org
Tue Oct 6 18:43:30 PDT 2009
On Tue, Oct 6, 2009 at 10:26 AM, Andrew Sullivan <ajs at crankycanuck.ca> wrote:
> I'm tempted to note that, if even one of the principal developers has
> to couch his remark that way, we maybe need some test case proofs one
> way or the other.  Vick, care to oblige?

I'm of the opinion that if the implicit trigger still fires on the
replica it is not a horrible thing.  It just wastes some cycles.  In
my case ultimately the new table was to be added to replication.  I
just wanted to avoid the burden of shutting down things enough that an
EXECUTE SCRIPT could successfully run.  The rate of change on this DB
is very high and I usually end up with some failure unless I disable
sections of the app.


More information about the Slony1-general mailing list