A simple restart of all slon processes seems to have resolved the issue. 
weird ....

Cyril Scetbon wrote:
> I use test_slony_state and see some informations like :
>
> Check of event info
> ---------------------------------------------------
> Problem : Events not propagating to node 2
> Problem : Events not propagating to node 4
> Problem : Events not propagating to node 3
>
> Check of sl_confirm aging
> ---------------------------------------------------
> Confirmations not propagating from 2 to 1
> Confirmations not propagating from 2 to 3
> Confirmations not propagating from 2 to 4
> Confirmations not propagating from 3 to 1
> Confirmations not propagating from 3 to 2
> Confirmations not propagating from 3 to 4
> Confirmations not propagating from 4 to 1
> Confirmations not propagating from 4 to 2
> Confirmations not propagating from 4 to 3
>
> You can see the results on one of my databases :
>
> - for Confirmations
>
> select con_origin, con_received, min(con_seqno) as minseq, 
> max(con_seqno) as maxseq, date_trunc('minutes', 
> min(now()-con_timestamp)) as age1, date_trunc('minutes', 
> max(now()-con_timestamp)) as age2, min(now() - con_timestamp) > 
> '00:30:00' as tooold from _pns_slony_voila_archi_0.sl_confirm group by 
> con_origin, con_received order by con_origin, con_received;
> con_origin | con_received | minseq  | maxseq  |   age1   |   age2   | 
> tooold 
> ------------+--------------+---------+---------+----------+----------+-------- 
>
>        101 |          102 | 4464029 | 4464792 | 00:00:00 | 00:16:00 | f
>        101 |          103 | 4464027 | 4464792 | 00:00:00 | 00:16:00 | f
>        101 |          104 | 4464024 | 4464792 | 00:00:00 | 00:16:00 | f
>        102 |          101 |      29 |      29 | 03:39:00 | 03:39:00 | t
>        102 |          103 |      29 |      29 | 03:39:00 | 03:39:00 | t
>        102 |          104 |      29 |      29 | 03:39:00 | 03:39:00 | t
>        103 |          101 |      62 |      62 | 03:39:00 | 03:39:00 | t
>        103 |          102 |      62 |      62 | 03:39:00 | 03:39:00 | t
>        103 |          104 |      62 |      62 | 03:39:00 | 03:39:00 | t
>        104 |          101 |      57 |      57 | 03:39:00 | 03:39:00 | t
>        104 |          102 |      57 |      57 | 03:39:00 | 03:39:00 | t
>        104 |          103 |      57 |      57 | 03:39:00 | 03:39:00 | t
>
>
> - for Events
>
> select ev_origin, min(ev_seqno), max(ev_seqno),
>         date_trunc('minutes', min(now() - ev_timestamp)),
>         date_trunc('minutes', max(now() - ev_timestamp)),
>         min(now() - ev_timestamp) > '00:30:00' as agehi
>     from _pns_slony_voila_archi_0.sl_event group by ev_origin;
> ev_origin |   min   |   max   | date_trunc | date_trunc | agehi
> -----------+---------+---------+------------+------------+-------
>       103 |      62 |      62 | 03:49:00   | 03:49:00   | t
>       104 |      57 |      57 | 03:49:00   | 03:49:00   | t
>       102 |      29 |      29 | 03:49:00   | 03:49:00   | t
>       101 | 4464493 | 4465346 | 00:00:00   | 00:14:00   | f
>
>
> What can be the source of these errors and how to track them ? FYI, I 
> have logs but in debug level 1.
>
> I did not have issues in 1.2.15.
>
> Regards.

-- 
Cyril SCETBON