Hello Bruce. I'd be willing to bet that if a firewall
is involved, it's been especially tuned for Oracle, whereas special rules
weren't set for Postgres. As soon as those rules get put in, things will
probably be very equivalent.
-Brian
On Mar 2, 10:15am, Bruce Badger wrote:
} Subject: Re: Beating Oracle
} Bruce Momjian wrote:
}
} >Bruce Badger wrote:
} >
} >>Tom Lane wrote:
} >>
} >>>Bruce Badger <bbadger(at)BadgerSE(dot)com> writes:
} >>>
} >>>>Due to network glitch, a PostgreSQL connection over IP is being dropped.
} >>>>
} >>>Perhaps there is something wrong with your TCP stacks. I do not think
} >>>it's Postgres' responsibility to second-guess the network transport
} >>>code...
} >>>
} >>> regards, tom lane
} >>>
} >>Not *my* stack, BTW, but that aside I agree with you. The people
} >>involved suspect that a firewall is causing the problem.
} >>
} >>The thing is that from their point of view, Oracle keeps going where
} >>PostgreSQL does not. Given that both are running over the same network,
} >>the response of these people could well be to simply move everything to
} >>Oracle.
} >>
} >>The question, then, is: what can we PostgreSQL people can do to match
} >>this Oracle ability??
} >>
} >
} >Unless there is more information, nothing. This is the first time I
} >have heard of this in my 6 years with PostgreSQL.
} >
} OK - now I have spoken with you sages, I will indeed go back and see if
} these people have any more information.
}
} Many thanks.
}
}
}
} ---------------------------(end of broadcast)---------------------------
} TIP 4: Don't 'kill -9' the postmaster
>-- End of excerpt from Bruce Badger