Re: [wsrm] issue with use of SOAP Faults

[Tom Rutt <tom@coastin.com>]

Comments inline

Tom Rutt

Jacques Durand wrote:

> In section 4.5 the requirement to use SOAP Faults in specific cases,
> may not be feasible:
>
> Because the RMP is unable to distinguish a message that belongs to a 
> Req-Resp MEP, from one
> taht belongs to a One-Way, it is not clear how the RMP can behave 
> differently based on this:
> - One-Way should not return SOAP Faults in HTTP response (to be WS-I 
> 1.0 compliant)
> - but Req-Resp are required to do so in case below:
>
>             "In case of a Request-Response WSDL operation type, when
>             the message cannot be passed to the consumer due to a
>             failure in processing the RM headers, and therefore no
>             application response can be returned, a SOAP Fault MUST be
>             returned. If the RM Fault is a Message Format fault, a
>             SOAP client fault MUST be returned. If it is a Message
>             Processing fault, a soap:server fault MUST be returned.
>
>             The latter case also applies to responses to duplicate
>             messages that are not delivered. In case a "Response"
>             ReplyPattern was required, the RM-Reply MUST be returned
>             in the header of the SOAP Fault message..."
>
> The intent here , if I remember, was to make sure that the lack of 
> application response would be escalated properly as a fault / 
> exception to the Sender app (or "producer"), which would be achieved 
> by adding a SOAP Fault in the HTTP response.
>
Since there is no return message information , and the sender is 
expecting a response in the soap body,
the soap fault is the only way to give a valid return for this case.

> One way out of this would be to link the use of SOAP Faults in HTTP 
> responses to the
> replypattern "Response", instead of to the WSDL op type
> (The response reply pattern is supposed to be used only for WSDL 
> Req-Resp ops.)
>
I think that is how I read the text above, it is the response reply 
pattern which triggers the production
of the soap fault in this case.

> Also,  in case replypattern is something else: do we still need to use 
> SOAP Faults? (I don't
> believe there is a point to it.)
>
I do not think there is any text in the current document which implies 
this.  I agree, we will not send soap
faults for callback or poll response.

> Jacques
>
>

-- 
----------------------------------------------------
Tom Rutt	email: tom@coastin.com; trutt@us.fujitsu.com
Tel: +1 732 801 5744          Fax: +1 732 774 5133