On 3 January 2012 19:29, Steve Singer <ssinger at ca.afilias.info> wrote:
> On 12-01-03 12:31 PM, Martijn van Oosterhout wrote:
>> Sorry, I should have mentioned: slony-I 2.0.7, PostgreSQL 9.0.3
> From your log snippet I would say either
>
> 1) You are not subscribed to the set, and node 2 has not processed the
> ENABLE_SUBSCRIPTION event.
>
> 2) You are subscribed to the set but are not receiving or processing any
> events from node 1 (for some unknown reason)
>
> Querying sl_status on the master would have told you how far behind the
> slave is.

sl_status was fine, that's what the nagios check does and it said
everything was fine.

But, reviewing the script now (based on the psql_replication_check.pl)
it looks like it says green also when there is no active replication.
So I think the simplest explanation is the subscribe wasn't done
properly.

Thanks for the help,
-- 
Martijn van Oosterhout <kleptog at gmail.com> http://svana.org/kleptog/