Hi! Have you had a look to pgpool logs? Did you execute pgpool in 'debug' mode? It could be that somewhere in time pgpool finds mismatch between those databases thus degenerating the slaves (if "replication_stop_on_mismatch = true"). Anyway, you will see it on logs.
2009/8/11 Joe Conway <[email protected]> > Hello all, > > I'm a pgpool newbie with a question or two. > > I'm trying to test pgpool for an existing application. What I did was to > dump the existing database from one Postgres cluster, and load it into > another Postgres cluster running on a different port of the same > machine. One of the Postgres clusters was installed from RPM. The other > was compiled from source. Both run perfectly on their own. > > backend_hostname0 = 'localhost' > backend_port0 = 5432 > backend_weight0 = 1 > backend_data_directory0 = '/var/lib/pgsql/data' > backend_hostname1 = 'localhost' > backend_port1 = 55439 > backend_weight1 = 1 > backend_data_directory1 = '/usr/local/pgsql-8.3/data' > > I defined one partitioned table and two replicated tables in the pgpool. > There are many other tables. > > At that point I tried to use psql to run a query on the partitioned > table. I never get that far. Instead I get this: > > psql -p 9999 mydatabase > psql: server closed the connection unexpectedly > This probably means the server terminated abnormally > before or while processing the request. > > However, if I modify the backend definition so that both databases point > to the same postgres cluster, the psql login succeeds. > > What am I doing wrong? > > Thanks, > > Joe > > > > _______________________________________________ > Pgpool-general mailing list > [email protected] > http://pgfoundry.org/mailman/listinfo/pgpool-general > >
_______________________________________________ Pgpool-general mailing list [email protected] http://pgfoundry.org/mailman/listinfo/pgpool-general
