I yet have error  in slony, below a show my erro. My unique key is
"chave_idx" and my table in Slony is "trusted".

- Screen in the Slave:
2006-09-14 02:00:47 BRT DEBUG1 cleanupThread:    0.056 seconds for
cleanupEvent()
2006-09-14 02:00:47 BRT DEBUG1 cleanupThread:    0.027 seconds for delete
logs

- Screen in the master:
2006-09-14 13:31:56 BRT DEBUG1 cleanupThread:    0.134 seconds for
cleanupEvent()
2006-09-14 13:31:56 BRT DEBUG1 cleanupThread:    0.060 seconds for delete
logs
2006-09-14 13:42:52 BRT DEBUG1 cleanupThread:    1.139 seconds for
cleanupEvent()
2006-09-14 13:42:53 BRT DEBUG1 cleanupThread:    0.300 seconds for delete
logs

- Log (/var/log/postgresql/postgres.log) in the master:
2006-09-14 14:08:35 [11066] ERROR:  duplicate key violates unique constraint
"trusted_chave_key"


 Does someone has idea about this?

 After 2 hours my Slony master stop to work, nad I need restart slony
master.


2006/9/14, cbbrowne at ca.afilias.info <cbbrowne at ca.afilias.info>:

> >  I have the table "trusted" that slony is working (replicat). When I
> exec
> > daemon slony I receive this message:
> >
> > 2006-09-14 10:06:18 BRT ERROR  cannot get sl_local_node_id - ERROR:
> > schema
> > "_trusted" does not exist
> > 2006-09-14 10:06:18 BRT FATAL  main: Node is not initialized properly
> >
> >
> >  How can I do to Slony never show me this message? Could someone help
> me?
> >
> >  I look that schema "_trusted"  exist, and slony is working. But after 1
> > day
> > I need to kill and exec daemon slony, after this Slony return to work.
>
> The error message indicates that the _trusted schema doesn't exist, so
> that the database being accessed does not contain any of the Slony-I
> configuration tables.
>
> I can see two ways for that to be the case after you had replication
> "working":
>
> 1.  Is it possible that the slon process is connecting to the wrong
> database?  Perhaps there are environment variables (PGHOST, PGPORT,
> PGDATABASE, ...) set that are redirecting the slon to the wrong place.
>
> You might turn connection logging on on your favorite databases, and start
> up the slon; if there's a connection to a WRONG place, that'll help
> diagnose what's up.
>
> 2.  The other possibility is that someone dropped the _trusted schema.
>
> That destroys the node, as far as Slony-I is concerned...
>
> The only way Slony-I would do this is if you submit either a DROP NODE or
> UNINSTALL NODE request via a slonik script...
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://gborg.postgresql.org/pipermail/slony1-general/attachments/20060914/43cb24f3/attachment.html