[Pgcluster-general] PgCluster: Second node in cluster does not work alone.

[indra wardhana]

Lia Domide wrote:
>
> Hello to everybody,
>
> I am new in using PostgreSQL, and I find PgCluster ides more than 
> interesting, but I encounter a problem while using it.
>
> I am using PgCluster 1.7.0rc7, PostgreSQL8.2.5 and Ubuntu 7.04
>
> I have two nodes configured in a standard schema:
>
> ------------ ---------------
>
> | pg1 | | pg2 |
>
> | rep1 | | rep2 |
>
> | lb1 | | lb2 |
>
> Because “ssh – 2” is used, I have configured/tested ssh connection 
> from each node without a password. The restore procedure (-R) from any 
> node is working fine.
>
> The configuration files are similar, except for the normal differences 
> (node name specification), and are attached to this message.
>
> The replication seems to work correctly, from each node to the other, 
> but I have a problem when I want to test the load balance services.
>
> 1. The first testing scenario works fine: pg2 (DB service) is stopped; 
> I can access, through JBDC, any of the load balancers.
>
> 2. The second test fails: with pg1 (DB service) stopped, when I want 
> to use lb2, this says: /Connection rejected: pglb could not connect to 
> server: no cluster available/ (as if the pg2 would be stopped too).
>
> What is strange is that, when restarting lb2 (pg1 still down), one 
> JDBC call to lb2 works fine, but the second try fails again 
> (“/Connection rejected: Sorry, backend connection is full”/).
>
> 3. Same message:/Connection rejected: Sorry, backend connection is 
> full / is received, when I start pg2, rep2, and lb2 and try to connect 
> through JDBC directly to lb2 (first try is working correctly and the 
> second throws the above error message).
>
> Could you give some hints why this is happening?
>
> Thanks in advance,
>
> Lia Domide.
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Pgcluster-general mailing list
> Pgcluster-general at pgfoundry.org
> http://pgfoundry.org/mailman/listinfo/pgcluster-general
>   
are you already declare pg1 and rep1 on /etc/hosts ?

=indra=