OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

ws-tx message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [ws-tx] Issue 084 - WS-AT: Completion protocol diagram shouldinclude coordinator-generated aborted transition


I believe this should be discussed after Issue 077 - WS-AT: Make completion protocol request-reply, permitting anon.

I also think that we need to consider how to report an indeterminate transaction state via the CP (result of a 2PC proto error). Part of issue 041.

Per se, the proposed change is OK as a reflection of the new state table.

However, I am concerned that the implications of this innovation were barely discussed in resolving 056, particularly as this affects interoperable interactions.

Specifically, interop scenario 4.2 (early abort) will now engender a spontaneous response to the CP participant. Hitherto, a likely scenario for an interop implementer is to always register a CP participant as part of standard start up, in case a commit/abort decision needs to be communicated.

Success criterion 3 is transaction aborted. Easy way to ascertain outcome is to use CP participant in IA to attempt Commit, and read out the CP response message.

This change affects that approach, implying that implementations must implement the spontaneous notification, or decide to treat it as optional, or use some local determination of outcome status. Note that the sending of such a notification will move state to None, which will in turn be likely to affect ability to determine local status (impls that don't remember outcome will now move into mode of returning Unknown Transaction or local API equiv if an enquiry is made to determine the outcome).

This is late breaking change that affects interoperable behaviour, implementations, interop scenario suites, and has received very little attention from the TC.

Alastair


Ram Jeyaraman wrote:

Any objections to the proposed resolution?

 

If there are no objections, I request the chairs to kindly add this to the list of issues for the next TC call.

 

From: Ram Jeyaraman [mailto:Ram.Jeyaraman@microsoft.com]
Sent: Monday, July 17, 2006 4:37 PM
To: ws-tx@lists.oasis-open.org
Subject: [ws-tx] Issue 084 - WS-AT: Completion protocol diagram should include coordinator-generated aborted transition

 

This is identified as WS-TX issue 084.

 

Please ensure follow-ups have a subject line starting "Issue 084 – WS-AT: Completion protocol diagram should include coordinator-generated aborted transition".

 

Protocol:  WS-AT
 
Artifact:  spec
 
Draft:  AT spec cd 2
 
Link to the document referenced:
 
http://www.oasis-open.org/committees/download.php/18889/wstx-wsat-1.1-spec-cd-02.doc

 

Section and PDF line number:  Completion protocol diagram in lines 160-161.

 

Issue type:  Editorial

 

Related issues:
 
Issue 056
 
Issue Description:

Consistent with resolution to issue 56, the Completion protocol diagram should show the coordinator generated aborted transition from Active to Ended state.

 
Proposed resolution

 

In the Completion protocol diagram add a coordinator-generated transition from Active to Ended state. Label the transition “Aborted”.



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]