Re: [Sigtran] I-D ACTION:draft-ietf-sigtran-m3ua-implementors-guide-02.txt

Tolga,

Tolga Asveren wrote:                 (Sat, 21 Dec 2002 15:44:00)
> [TOLGA] On that thread, also my intial opinion was
> that ASPIA-ACK is the most appropriate action from
> SGP, when NIF is not available. But after some time I
> noticed the scenario, where SGP routes between two
> ASPs and also to/from TDM domain SS7 nodes. In such a
> case, one could argue that unavailability of NIF might
> not hinder any message transfer activity on SGP
> between two ASPs. So, I thought, DUNA for the relevant
> destinations (destinations in TDM domain) might be a
> better alternative, so that SGP can continue to route
> between ASPs. But now, what I think is, all of this is
> too implementation dependent (what would one expect,
> if NIF is involved?). Depending on implementation, one
> could have NIF involved also for message routing
> between ASPs, or not. So, after all of this, I again
> return to my very initial proposal : There is no need
> to say anything about unavailability of NIF. There are
> too many implementation dependednt things (whether it
> migth be unavailable from a single SGP but available
> from the others or is on/off for the whole SG, whether
> it playa a role in routing between ASPs etc...) And if
> the consensus is that there is a need to describe the
> behaviour in such a case, then I would think, it is
> not a very good idea to use "MUST" for any described
> activity. Depending on implementation/configuration
> both DUNA and ASPIA-ACK might be fine and even I
> wouldn't choose it I still think even dropping
> association won't be wrong.
> 

I agree, how a SGP chooses to report this situation is
implementation depedent.  The SG implementation has all
three procedures at its disposal, the results of which
are well defined.  It is up to the SG designer to choose
the approach most appropriate for her implementation.

I would have no problem if the section discussed the
implications of each procedure, without recommending
(or requiring) one or the other.

--brian

-- 
Brian F. G. Bidulock
bidulock@openss7.org
http://www.openss7.org/
_______________________________________________
Sigtran mailing list
Sigtran@ietf.org
https://www1.ietf.org/mailman/listinfo/sigtran