I have to db clusters on 1 server and 2 other ha-clusters (also with 2 db
clusters) spread over the network.
Since I have these sl_nodelock messages also in my 2 local postgres logs I
don't think that this is caused by a network problem.

Can I somehow enable more logging to analyse this in more detail?

Thx
felix

-----Original Message-----
From: chris [mailto:cbbrowne at ca.afilias.info] 
Sent: Mittwoch, 16. Juli 2008 18:52
To: LINHART Felix
Cc: slony1-general at lists.slony.info
Subject: Re: [Slony1-general] sl_nodelock messages

"LINHART Felix" <Felix.LINHART at frequentis.com> writes:
> I have just noticed that my postgres logfile receives continuously 
> messages from slony listed below::p>
>
> :p>
>
> 2008-07-16 12:37:50 CEST NOTICE:  Slony-I: cleanup stale sl_nodelock 
> entry for pid=5810:p>
>
> 2008-07-16 12:37:50 CEST CONTEXT:  SQL statement "SELECT  
> "_slony1".cleanupNodelock()":p>
>
>         PL/pgSQL function "cleanupevent" line 77 at PERFORM:p>
>
> 2008-07-16 12:51:35 CEST LOG:  could not receive data from client: 
> Connection timed out:p>
>
> 2008-07-16 12:51:35 CEST LOG:  unexpected EOF on client connection:p>

Since these are "NOTICE" log records, it does not indicate any error as far
as Slony-I is concerned.

When the slon falls over, and another starts up, the new one has to clean up
after the old one.  It's not a real big deal, though it suggests that you
have some sort of network problem or something that is causing slon processes
to fall over so they need to be restarted.

That falls outside what Slony-I is, itself, doing.
--
select 'cbbrowne' || '@' || 'linuxfinances.info';
http://cbbrowne.com/info/lsf.html Rules  of the  Evil Overlord  #145. "My
dungeon cell  decor  will not feature exposed pipes.  While they add to the
gloomy atmosphere, they are good  conductors of vibrations and  a lot of
prisoners know Morse code." <http://www.eviloverlord.com/>.