RE: [wsrf] ISSUE: Should resource lifetime require WS-Resource.

["Sedukhin, Igor S" <Igor.Sedukhin@ca.com>]

That is right. Application is to WS-Resouces only. That is, an EPR
without reference properties may still work :), but the one which refers
to that WS-Resource would not. There is no assumptions made about EPRs
refering to non WS-Resources here.

I used EPR to describe this, but the actual semantics is "don't talk to
here no more -- nobody home" :)...

So normatively it may be descibed as follows "Destroy messge exchange
results in the WS-Resource being destroyed and not available for access
after that. Any means used previously to access the WS-Resource must
result in a fault." So that is my suggestion.


-- Igor Sedukhin .. (igor.sedukhin@ca.com)
-- (631) 342-4325 .. 1 CA Plaza, Islandia, NY 11749


-----Original Message-----
From: Jeffrey Frey [mailto:jafrey@us.ibm.com] 
Sent: Monday, November 01, 2004 1:45 PM
To: Sedukhin, Igor S
Cc: meder@mcs.anl.gov; Steve Graham; wsrf-oasis
Subject: RE: [wsrf] ISSUE: Should resource lifetime require WS-Resource.





I guess there is no hard architectural reason to limit this to a
WS-Resource. But the intent was clearly to destroy the specific instance
of the stateful entity represented by the WS-Resource. I find it
confusing to destroy a "stateless" web service, although as you point
out it might simply mean that the service stop accepting requests. I
assume this would render a deployed "stateless" web useless. I guess I
do not like the application of this semantic to non WS-Resources. This
was about instance lifecycle.

And to be precise, we don't send things to an EPR. We send messages to a
web service addressed by one or more EPRs. The operational state of the
WS-Resource/Web service being destroyed is the issue at hand here,
regardless of which of the many possible EPRs may have been used to
address it.


Jeffrey Frey

IBM Distinguished Engineer
OnDemand System Architecture and Design
Phone: 845-435-3067  Tie: 8-295-3067  Cell: 914-456-6556
Notes: Jeffrey Frey/Poughkeepsie/IBM@IBMUS
Internet: jafrey@us.ibm.com




 

             "Sedukhin, Igor

             S"

             <Igor.Sedukhin@ca
To 
             .com>                     Steve Graham/Raleigh/IBM@IBMUS,

                                       <meder@mcs.anl.gov>

             11/01/2004 12:28
cc 
             PM                        "wsrf-oasis"

                                       <wsrf@lists.oasis-open.org>

 
Subject 
                                       RE: [wsrf] ISSUE: Should resource

                                       lifetime require WS-Resource.

 

 

 

 

 

 





The WS-Resource is destroyed. That is, don't send nothing to its EPR no
more -- there is nobody there. I think that this is more obvious
semantics than destruction of a "stateful resource component of the
WS-Resource"
which, by the way, has never been normatively defined anywhere else.



-- Igor Sedukhin .. (igor.sedukhin@ca.com)
-- (631) 342-4325 .. 1 CA Plaza, Islandia, NY 11749




From: Steve Graham [mailto:sggraham@us.ibm.com]
Sent: Monday, November 01, 2004 8:37 AM
To: meder@mcs.anl.gov
Cc: wsrf-oasis
Subject: Re: [wsrf] ISSUE: Should resource lifetime require WS-Resource.


Sam:
This is a good question.

Now, what is the semantic of Destroy() if no WS-Resource is assumed.
Currently, the text from WSRF-RL version 1.2 draft 03 reads:

224 If the WS-Resource accepts the Destroy request message, upon receipt
of this message the WS-
225 Resource MUST either (1) destroy the implied stateful resource
component of the WS-Resource
226 and return the following DestroyResponse message to acknowledge
successful destruction, or
227 (2) return a fault message indicating failure. Note that the
destruction of the stateful resource
228 component of the WS-Resource effectively destroys the WS-Resource.

So, if we allow the case where the Destroy is sent to something other
than a WS-Resource, what is the semantic?  Ie what can the requestor
imagine is destroyed in successful response to the request?

sgg
++++++++
Steve Graham
(919)254-0615 (T/L 444)
STSM, On Demand Architecture
Member, IBM Academy of Technology
<Soli Deo Gloria/>
++++++++


 

 Samuel Meder

 <meder@mcs.anl.gov>

 

 
To 
 10/29/2004 09:41 PM                         wsrf-oasis

                                             <wsrf@lists.oasis-open.org>

 
cc 
       Please respond to

             meder
Subject 
                                             [wsrf] ISSUE: Should
resource 
                                             lifetime require
WS-Resource. 
 

 

 

 

 

 

 






Specifically should the destroy operation require that the Web Service
implementing it is also a WS-Resource.

/Sam