Steve Singer ssinger at ca.afilias.info
Wed Apr 30 19:16:28 PDT 2014
On 04/30/2014 12:36 PM, Glyn Astill wrote:
>> From: Steve Singer <ssinger at ca.afilias.info>
>
>> To: Glyn Astill <glynastill at yahoo.co.uk>; "slony1-general at lists.slony.info" <slony1-general at lists.slony.info>
>> Sent: Wednesday, 30 April 2014, 16:57
>> Subject: Re: [Slony1-general] Slony 2.2 failover changes
>>
>>
>> On 04/28/2014 09:56 AM, Glyn Astill wrote:
>>
>> This is sounding like a bug, what is your path network?
>> Do you have paths between all nodes in both directions or something else?
>>
>
> Yeah I have paths between them all in both directions:
>
> TEST=# select * from _test_replication.sl_path order by pa_server, pa_client;
>   pa_server | pa_client |                   pa_conninfo                   | pa_connretry
> -----------+-----------+-------------------------------------------------+--------------
>           1 |         2 | dbname=TEST host=localhost port=5432 user=slony |           10
>           1 |         3 | dbname=TEST host=localhost port=5432 user=slony |           10
>           1 |         4 | dbname=TEST host=localhost port=5432 user=slony |           10
>           2 |         1 | dbname=TEST host=localhost port=5433 user=slony |           10
>           2 |         3 | dbname=TEST host=localhost port=5433 user=slony |           10
>           2 |         4 | dbname=TEST host=localhost port=5433 user=slony |           10
>           3 |         1 | dbname=TEST host=localhost port=5434 user=slony |           10
>           3 |         2 | dbname=TEST host=localhost port=5434 user=slony |           10
>           3 |         4 | dbname=TEST host=localhost port=5434 user=slony |           10
>           4 |         1 | dbname=TEST host=localhost port=5435 user=slony |           10
>           4 |         2 | dbname=TEST host=localhost port=5435 user=slony |           10
>           4 |         3 | dbname=TEST host=localhost port=5435 user=slony |           10
> (12 rows
>


Can you send me the slon logs for slon4 ? Ideally at debug level 4.
Also when the cluster is in this state: on node 4 do a
select * FROM _test_replication.sl_node;

and

select * FROM _test_replication.sl_listen where
li_origin=1, li_receiver=4;



>> Does it happen everytime you test or only sometimes?
>
> Yeah, happens every time.
>
> Glyn
>



More information about the Slony1-general mailing list