Hiroshi Inoue wrote:
(B>
(B> > -----Original Message-----
(B> > From: Tom Lane
(B> >
(B> > Bruce Momjian <[EMAIL PROTECTED]> writes:
(B> > > Tom Lane wrote:
(B> > >> You're not considering the possibility of a transient communication
(B> > >> failure.
(B> >
(B> > > Can't the master re-send the request after a timeout?
(B> >
(B> > Not "it can", but "it has to".
(B>
(B> Why ?$B!!(BMainly the coordinator(slave) not the participant(master)
(B> has the resposibilty to resolve the in-doubt transaction.
(B
(BAs far as I see, it's the above point which prevents the
(Badvance of this topic and the issue must be solved ASAP.
(B
(BAs opposed to your answer
(B Not "it can", but "it has to",
(Bmy answer is
(B Yes "it can", but "it doesn't have to".
(B
(BThe simplest senario(though there could be varations) is
(B
(B[At participant(master)'s side]
(B Because the commit operations is done, does nothing.
(B
(B[At coordinator(slave)' side]
(B 1) After a while
(B 2) re-establish the communication path between the
(B partcipant(master)'s TM.
(B 3) resend the "commit requeset" to the participant's TM.
(B 1)2)3) would be repeated until the coordinator receives
(B the "commit ok" message from the partcipant.
(B
(BIf there's no objection from you, I would assume I'm right.
(BPlease don't dodge my question this time.
(B
(Bregards,
(BHiroshi Inoue
(B http://www.geocities.jp/inocchichichi/psqlodbc/
(B
(B---------------------------(end of broadcast)---------------------------
(BTIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
