Re: Recommendation for SUA modifications

Lincoln,

Haresign Lincoln wrote:                                                           (Thu, 13 Oct 2005 14:03:03)
> Brian,
> 
> What the SG does is implementation dependent.  I'm not sure we should
> specify it in the requirement.  Today, the RFC says:
> 
>    The SUA layer at an ASP or IPSP MAY indicate local congestion to an
>    SUA peer with an SCON message.  When an SG receives a congestion
>    message (SCON) from an ASP, and the SG determines that an endpoint is
>    now encountering congestion, it MAY trigger congestion procedures of
>    the relevant SCCP standard. 

I am familiar with the passage, I was largely responsible for it
being included in SUA, I wrote the text for M3UA, and was
responsible for the MAY.

Again, we put "MAY trigger" because the congestion procedures of
different National SCCP standards differ.  We put "MAY indicate"
because it is easy to design an SG that does not need the indication
to properly effect SCCP congestion procedures.

If you are implementing for 3GPP, ETSI TS 102 143 is applicable and
says:

 "When an SG receives a congestion message (SCON) from an ASP, and
 the SG determines that an endpoint is now encountering congestion,
 it shall trigger congestion procedures as per ITU-T Recommendation
 Q.714[4]."

No MAY.  But that is fine because ETSI TS 102 143 also restricts the
relevant SCCP protocol to the ETSI modified versions that support
SCCP congestion (but not subsystem congestion).  ETSI TS 102 143
also nicely affirms the intended use of SCON from ASP to SG.

> 
> Note the MAY.  In our implementation, we also take actions such as
> diverting traffic to non congested ASPs.  But in any case, the RFC
> ALLOWS the SCON from ASP to SG.  What the SG wants to do with it is up
> the the SG.

Well, no, not if you are deploying 3GPP in accordance with ETSI
specifications: then, the SG's actions appear to be dictated by
ETSI TS 102 143.

As the ETSI document follows more formal standards body approaches,
I don't belive that it allows an SUA SG in the 3GPP network to do
whatever it wants with the message, unless of course the
implementation is completely internal to an operator's network (in
which case they can do as they please anyway) or there exists a
bilateral agreement to that effect between adminstrative domains
(operators).

--brian

-- 
Brian F. G. Bidulock
[email protected]
http://www.openss7.org/