On January 6, 2005 11:20 am, Christopher Browne wrote:
> Brad Nicholson wrote:
> > Vivek Khera wrote:
> >> On Jan 5, 2005, at 1:03 PM, sarlav kumar wrote:
> >>> I do the following things:
> >>>
> >>> pg_dump -s -h $MASTERHOST -U $REPLICATIONUSER $MASTERDBNAME > dumpfile
> >>> I am attaching the dumpfile with this email. If you look at the
> >>> dumpfile, it creates an operator "=" though I did not create the
> >>> operator in the database.
> >>
> >> you're dumping a schema that's been altered by slony already.  Jan
> >> posted a script here the other day to get a "clean" dump from a
> >> database that's already slony-fied.
> >
> > You could also import the dump, and then drop the slony schema with
> > the CASACADE option to clean out the slony stuff.
>
> I should point out one problem with the "drop schema" scenario; that
> doesn't cope with tables where Slony-I added in a column to provide a
> primary key.  Those columns are not in the Slony-I namespace, so they
> don't get dropped.
>
> And they reference a sequence that gets dropped, so that things wind up
> not quite totally clean :-(.

The upcoming Slony-I version 1.1 has a slonik command (REPAIR CONFIG) you can 
run that will "clean up" a db that has been dump/restored to allow you to 
continure replicating, or at least perform other administrative tasks ie drop 
node on it.

> _______________________________________________
> Slony1-general mailing list
> Slony1-general at gborg.postgresql.org
> http://gborg.postgresql.org/mailman/listinfo/slony1-general

-- 
Darcy Buskermolen
Wavefire Technologies Corp.
ph: 250.717.0200
fx:  250.763.1759
http://www.wavefire.com