AFAIK, Slony says is mandatory to use some customized scripts to create tables on a cluster, due to sync among nodes.
research the documentation, I'm pretty sure that this issue is related to the way slony do this. On Thu, Jul 30, 2009 at 6:18 AM, Nimesh Satam<[email protected]> wrote: > Hi, > > We are using pgpool-2.1 in: > > Load balance and master slave mode. Slony is used for the purpose of > replication. Connection_cache is set to off. > > We have been facing problem with Temp table creation while using pgpool > > 1) CREATE TEMP TABLE tmp_met AS SELECT * from x.x; > > 2) CREATE INDEX tmp_met_indx ON tmp_met(id); > > 3) CREATE TEMP TABLE tmp_ch (key integer PRIMARY KEY); > > 4) INSERT INTO tmp_ch (key) SELECT key FROM rpt_ch WHERE id IN (135848, > 135841); > > Queries from 1 – 3 are seen in data_logs on both the replication database > (master and slave) and are working fine. But the third temp table creation > query throws an error on the slave db with the temp table “tmp_ch” does not > exist. Shouldn’t this query be sent to the master database? > > All the queries are binded together in a single transaction and hence should > not be giving any errors. Can anybody pinpoint what’s going wrong here? > > Regards, > > Nimesh. > > _______________________________________________ Pgpool-general mailing list [email protected] http://pgfoundry.org/mailman/listinfo/pgpool-general
