On Mon, Oct 13, 2008 at 5:05 PM, Stuart Bishop <stuart at stuartbishop.net> wrote:

> So what was the final recommended process for building a stand alone
> database from a pg_dump of a replicated node?

> So if I'm reading this thread correctly, the alternative is 'DROP _sl
> CASCADE;', which doesn't do a full cleanup. Is there no supported
> disaster recovery procedure?

So to (hopefully) answer my own question, the following seems to Do
The Right Thing™:

pg_dump --oids --format=c --file=master.dump master_db
createdb staging_db
pg_restore -d staging_db master.dump
slonik << EOM
cluster name = sl;
node 1 admin conninfo = 'dbname=staging_db user=slony';
repair config (set id = 1, event node = 1, execute only on = 1);
repair config (set id = 2, event node = 1, execute only on = 1);
uninstall node (id = 1);
EOM

Can anyone who actually knows what they are doing confirm or ridicule
this recipe?

-- 
Stuart Bishop <stuart at stuartbishop.net>
http://www.stuartbishop.net/