On Monday November 1 2004 4:20, Andrew Sullivan wrote:
> On Mon, Nov 01, 2004 at 10:47:57AM +1100, Justin Clift wrote:
> > Hmmm, it would make more sense for Slony to internally track the
> > progress of each sequence, so it doesn't blindly do setval() each timem
> > wouldn't it?
>
> What does it mean "to internally track the progress of each
> sequence"?  Does this mean that when a sequence is updated, Slony
> somehow gets notified and remembers the value?

Not necessarily.  Slony is already polling all sequences at each sync 
interval.  A first-level improvement might be for slon to simply keep track 
of what each last_value was the last time we checked, and if its not 
different from *this* check, don't update the slave.  That would eliminate 
the redundant updates without any touching of nextval() functionality.

But it'd be nice to eliminate the polling altogether if possible.  I don't 
know if there is a better way, but if one could trigger off a sequence 
change, that would seem to remove the polling requirement as well as the 
internal tracking need.

Ed