[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: RE: [ws-rx] i075 new proposal
I would add two items to your list of justifications: 4) Communicating these operational parameters in the CreateSequenceResponse allows applications that are not using WSDL (for whatever reason) to determine their value. 5) Removing these parameters from WSRM-Policy eliminates the possibility of contention between different policies set on endpoints that share the same sequence. I don't think it is possible for endpoints to have conflicts about whether or not they support WSRM. If an endpoint doesn't support WSRM it can't be a part of any sequence so (from the perspective of WSRM) there is nothing to disagree about. I think we need to separate the idea of presenting ack interval and max message at runtime from the idea of removing them from WS-RM Policy. Logically there are four combinations: (1) in-policy, not-in-runtime: current spec (2) not-in-policy, in-runtime: this proposal (3) in-policy, in-runtime: Anish's proposal(?) (4) not-in-policy, not-in-runtime: Bob Freund(?) Do we have proposals for (3) and (4)? - g > -----Original Message----- > From: Paul Fremantle [mailto:paul@wso2.com] > Sent: Thursday, January 05, 2006 4:44 AM > To: wsrx > Subject: [ws-rx] i075 new proposal > > Based on the discussion at the F2F I believe that we should > remove AcknowledgementInterval and MaxMessageNumber from WSRM > Policy and make them parts of the CreateSequenceResponse message. > > The justification is this: > > 1) Firstly these are actually parameters of a sequence, not policy. > 2) The issue of which values are in use on a given sequence > is much clearer if they are associated with the CS/CSR > 3) It seems highly unlikely that these would be used in the > choice of a service. There might very very very occaisionally > be situations where a Source would reject the values. In that > case the overhead is marginally higher (a CS/CSR followed by > a terminate sequence versus looking at a Policy). However, I > think this scenario is so unlikely that it is irrelevant > compared to the benefits of simplification. > > I also believe that this helps shape up the dependency on the > WS-Policy > Framework: it means that none of the operational aspects of > the protocol are dependent on WS-Policy. The dependency is > only there if you need to find out whether a service supports > WSRM. (which by the way I *do* consider to be a "policy" in > the sense of the WS-Policy model). > > So as an amendment to the previous proposal (thanks Anish), > here is the new proposal. > > Modification to the WSRM-Policy spec > ----- > Section 2.1 add new text to follow line 109: > "When a RM Destination provides RM services for more than one > endpoint it is RECOMMENDED that all the endpoints should have > the same values for RM Policy parameters. If the RM Policies > are not the same then the RM Policy parameters in effect for > each Sequence is governed by the endpoint that was used for > the <wsrm:CreateSequence> message." > ----- > Remove lines 113-114 (AI and MM parameters) Remove lines > 125-134 (AI and MM parameters) Remove line 164 (AI example) > Remove text: "Line (13) indicates the RM Destination may > buffer acknowledgements for up to two-tenths of a second." > > ------ > Modifications to the WSRM spec: > After line 302 insert: > > <wsrm:AcknowledgementInterval Milliseconds="xs:unsignedLong" ... /> ? > <wsrm:MaxMessageNumber Number="xs:unsignedLong" ... /> ? > > After line 326 Insert: > /wsrm:CreateSequenceResponse/wsrm:AcknowledgementInterval > This element, if present, specifies the duration after which > the RM Destination will transmit an acknowledgement. If > omitted, there is no implied value. > /wsrm:CreateSequenceResponse/wsrm:AcknowledgementInterval/@Mil liseconds > The acknowledgement interval, specified in milliseconds. > /wsrm:CreateSequenceResponse/wsrm:MaxMessageNumber > This element, if present, specifies the maximum message > number that the RM Destination will accept. If omitted, the > default value of the maximum unsigned long will be assumed. > /wsrmp:RMAssertion/wsrm:MaxMessageNumber/@Number > The maximum message number. > ----- > > Paul > > -- > > Paul Fremantle > VP/Technology, WSO2 and OASIS WS-RX TC Co-chair > > http://bloglines.com/blog/paulfremantle > paul@wso2.com > > "Oxygenating the Web Service Platform", www.wso2.com > >
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]