we had similar problem but with JDBC. pgpool parallel mode does not support extended protocol (PREPARE/BIND ..) of this driver which is what I think you are rely on as well. with JDBC I'd get mirroring instead of partitioning, however when I do the same insert from the cli/psql - partitioning rules are applied properly and data is partitioned as expected one backend node only. my setup is two nodes running pgsql-8.3.3 single table inserts, partitioning on bigint. we had to re-consider using pgpool II because of this issue. V.
[EMAIL PROTECTED] wrote: > Hello again, > > >>> Hi. >>> >>> >>> >>>>>> It's now working in pgpool-II 2.1 RC1. No more problems. And yes I >>>>>> >>>> use >>>> >>>>>> PHP >>>>>> PDO and AdoDB. My previous post has more details. So what's the >>>>>> replicate_def table and replicate_select for? Thanks again. :) >>>>>> > > >>>> --- snipped --- >>>> > > Sorry. I take it back. It works with direct PDO INSERT statement (without > prepare, just execute the query string directly) but not with PDO > bindParam wherein I first PREPARE the statement and do multiple inserts > with a loop because I'm trying to migrate the records from MySQL to > PostgreSQL. Instead of partitioning the data, it copies the inserted row > to all nodes and when I browse the pgpool-II server, it displays 3 copies > of the row inserted! Can anyone explain to me what had just happened > please? Thanks. > > Regards, > Viril > > _______________________________________________ > 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
