On Thu, Jan 28, 2010 at 4:14 PM, Melvin Davidson <melvin6925 at yahoo.com> wrote:
>
> >what's your VERSION?!
>
> Tory I'm on Slony 1.2.14 with PG 8.2

Sorry Melvin, wasn't asking you , just correcting my protocol :)
>
>
>
> >Ya the problem is my slon cluster is not up at this time. It's a
> >script I run to setup the cluster, once it's setup from the master, I
> >start slon on the various nodes. So there will be no confirmation
> >until I start up slon on the nodes, and I generally don't do that
> >until the cluster is setup. So this may not be the solution for me. In
> >fact it's odd that it only complains about 1,2,4 and none of the
> >others.
>
> >So after adding the confirmed my script is stuck,  :) ya because slon
> >is not up, so I can't get a confirmation.
>
> >Tory
>
> Well Tory, you absolutely need slon up and running on the master and all subscribing nodes.  A subscribe causes all tables on the slave to be either truncated or deleted, then loaded with the data from the provided. But that can't happen without slon running!
>
>
Understood, but this works in a similar configuration now. The script
starts by dropping the slon schema, than sets paths, listeners, than
sets some other information (priviledges etc). Then I start slon on
all nodes and things start to replicate. I've never had slon running
when I run this prepare script. Just wondering if it's something
different now that I'm trying to use a slave as the provider for a
query server, vs having all servers hit masternode (origin) directly.
That could be the issue I guess.

Thank you sir!

Tory