RE: [ebxml-msg] Issue for your review

[David Elliot <david.elliot@xmlglobal.com>]

Hi Venkata,

Comments below...

>>> If we take the later route (which seems to be what people are
>>> saying) then the receiving MSH may get a HTTP 500 AFTER it has
>>> sent the MSH ack back.  Now the sender side thinks the receiver has
>>> got the message but in fact receiver side does not have it!??
>>
>>Essentially the Receiving MSH got the ebXML message, the Sending MSH
>>got the ebXML Ack, so delivery complete.  Whatever is reported by the
>>HTTP transport (5xx, 4xx, xxx) along the way does not affect the
>>outcome as the ebXML layer data was exchanged ok.
>
> I agree with you when the message has been persisted properly at
> the receiving side, but if the receiving side has been failed to
> persist then we still need to depend on the HTTP transport response.


If the Receiving MSH fails to persist or process the ebXML message then it
will not send an ebXML Ack.  So in this situation the Sending MSH would
not receive an ebXML Ack, hence after a given timeout would assume
delivery failure and retry the send (depending upon the ebXML reliability
settings).  Which means that the particulars of any HTTP response codes do
not enter the equation here...

Would you be able to lay out a scenario where the ebXML layer of a Sending
MSH must have details of a HTTP transport response to correctly perform an
ebXML message delivery ??

Best Regards,

Dave Elliot
XML Global Technologies