avelino escribió:
Hi.
I have the following configuration with pgcluster-1.9.0rc7:
- At host "cluster1" (192.168.200.196): one load balancer (port 5443),
one replication server (port 5445) and one cluster DB (port 5444)
- At host "cluster2" (192.168.200.197): another cluster DB (port 5444)
- At host "cluster3" (192.168.200.198): another cluster DB (port 5444)
I am executing pgreplicate and pglb into a "screen" session with
"-lnv" options to see all the log. All seems to be ok but when I try
to open a session from another machine with "psql" this is the result:
At client side:
[avel...@dulcinea ~]$ psql -h 192.168.200.196 -U dbadmin -p 5443 -l
psql: server closed the connection unexpectedly
This probably means the server terminated abnormally
before or while processing the request.
[avel...@dulcinea ~]$
At server side (pglb log):
2010-01-13 14:45:21 [3273] DEBUG:PGRscan_cluster:3 ClusterDB can be used
2010-01-13 14:45:21 [3348] DEBUG:I am 3348 accept fd 6
2010-01-13 14:45:21 [3273] DEBUG:PGRscan_cluster:192.168.200.196
[5444],useFlag->2 max->100 use_num->0
2010-01-13 14:45:21 [3348] DEBUG:Protocol Major: 1234 Minor: 5679
database: user:
2010-01-13 14:45:21 [3348] DEBUG:PGRdo_child(): SSLRequest: sent N;
retry startup
2010-01-13 14:45:21 [3365] DEBUG:I am 3365
But, when I connect with psql from the same machine it runs ok:
[dbad...@cluster1 ~]$ /opt/pgcluster-balancer/bin/psql -U dbadmin -h
192.168.200.196 -p 5443 -l
List of databases
Name | Owner | Encoding
-----------+---------+----------
postgres | dbadmin | LATIN9
template0 | dbadmin | LATIN9
template1 | dbadmin | LATIN9
(3 rows)
Help me, please. Thanks.
_______________________________________________
Pgcluster-general mailing list
[email protected]
http://pgfoundry.org/mailman/listinfo/pgcluster-general
I have realized that when I connect to load balancer at "cluster1" from
"cluster2" or "cluster3" the command psql runs ok:
[dbad...@cluster2 ~]$ /opt/pgcluster-cluster2/bin/psql -U dbadmin -h
cluster1 -p 5443 -l
List of databases
Name | Owner | Encoding
-----------+---------+----------
postgres | dbadmin | LATIN9
template0 | dbadmin | LATIN9
template1 | dbadmin | LATIN9
(3 rows)
[dbad...@cluster3 ~]$ /opt/pgcluster-cluster3/bin/psql -U dbadmin -h
cluster1 -p 5443 -l
List of databases
Name | Owner | Encoding
-----------+---------+----------
postgres | dbadmin | LATIN9
template0 | dbadmin | LATIN9
template1 | dbadmin | LATIN9
(3 rows)
The "psql" that fails is the one that comes with the PostgreSQL 8.3.8
(Fedora 10).
Please, help. Thanks in advance.
_______________________________________________
Pgcluster-general mailing list
[email protected]
http://pgfoundry.org/mailman/listinfo/pgcluster-general
