That's ridiculous. I haven't heard such a problem ever before. Can you tell me self contained test case? i.e. the table definition, SQL and pgpool.conf. -- Tatsuo Ishii SRA OSS, Inc. Japan
> Sure thing - > really nothing spectacular occurred. > > the table in the master node accepted the insert, and added a record, while > the other nodes did not. > the table has a SERIAL primary key, so it incremented while the other nodes > stayed, but got desynchronized for further inserts. > that's not a big deal though - this is testing, so it detached, recovered, > then attached those nodes to keep testing. > > if i need to - i can get away from SERIAL primary keys since that > convenience is outweighed by having solid failover and loadbalancing in my > mind. > > i'm going to go ahead and try the comment prior to selecting the function > and then report back. > > > > > > > i have also tried adding things like: > > > LOCK TABLE table_name IN SHARE ROW EXCLUSIVE MODE; > > > INSERT.... > > > > > > to no avail. > > > > Can you tell me more details on what happend? > > -- > > > > _______________________________________________ Pgpool-general mailing list [email protected] http://pgfoundry.org/mailman/listinfo/pgpool-general
