Sunil Kunisetty wrote:
>
> Patrick,
>
> I didn't cook up this requirement. I was just pointing to an existing
> one
> and my proposal was based on that.
>
> If it changes, I'll rework my proposal based on the new requirement.
>
> -Sunil
>
> Patrick Yee wrote:
>
>> IMHO, practically those persisted message IDs have to be cleaned up
>> regularly. Or else the persistence storage will be growing larger and
>> larger without limit. In fact, I think at some time a persisted
>> message ID can be cleaned up safely, depending on the message expire
>> time and the time need to complete all retries.
>
At the Redwood City meeting last May we came to general agreement that
the protocol
requires persistence to work (e.g, persistence of ids for duplicate
elimintation, persistence of
message contents for ordered delivery). However, some users might we
willing to accept
protocol failure under specific circumstances which are agreed witht the
system supplier (e.g, if the battery is not kept charged on a cell
phone, it might loose its "persistence" capability).
Thus this preagreed limit for persistence time is subject to user
contract, since it involves an
argreement as to what conditions are acceptable for tolerating protocol
failure.
I hope this helps.
Tom Rutt
WSRM Chair
>> Regards, -Patrick
>>
>>
>>
>>>> >> If someone asks for the status of a message that was successfully
>>>> >> passed on to the application layer or that expired and was purged,
>>>>
>>>>
>>>
>>> One of the current requirement reads "The receiver MUST persist the
>>> Message ID to support duplicate elimination". So to support this
>>> requirement, we anyway need to archive the ID.
>>>
>>>
>>>
>> To unsubscribe from this mailing list (and be removed from the roster
>> of the OASIS TC), go to
>> http://www.oasis-open.org/apps/org/workgroup/wsrm/members/leave_workgroup.php.
>
--
----------------------------------------------------
Tom Rutt email: ;
Tel: +1 732 801 5744 Fax: +1 732 774 5133
