Re: Congestion in M3UA/SUA and the use of the SCON

Haresign,

Haresign Lincoln wrote:                                                           (Mon, 10 Oct 2005 14:43:29)
> Brian,
> 
> Perhaps you were not aware, but additional GTT can also take place at an
> endpoint.

Yes, and the result of GTT is a point code.

> Many STPs will do GTT and they can allow further GTT to take
> place at the final destination (the signalling end point).

And to send the message to that end point requires a point code.

> In this
> case, it will arrive at an SG with a GTT in the called and calling
> address.

If there is no point code in the called address, it is taken from the
DPC of the MTP label.  SUA does not have an MTP label so this information
is included in the Called Address for SUA.

> The SG is not required to (and probably shouldn't) convert the
> called address to a PC/SSN.

For ASP operation it does.  SUA SG-ASP transport the SCCP-SCCP User
interface (see the RFC).  See Q.711.  CL and CREF are always delivered
to the SCCP User with a MTP-SAP (point code and SI value).

> So a CLDT may arrive at the ASP with JUST a GT in the called address.  

No.  There are some cases in Q.711 CO where the Responding Address can
differ from the Called Address during connection set up, but both still
contain an MTP-SAP.

> 
> And you may have designed your architecture so the ASPs have no need to
> know their point code (a much simpler configuration).  This is nothing
> NEW or outside the spec.

You can do that, however, to signal SCCP congestion requires identification
of the SCCP that is suffering congestion, and the SCCP is identified by its
MTP-SAP (point code and SI value), not by a Global Title.  Therefore, the
Affected Point Code is mandatory in the SCON, just as it is mandatory in
the N-PCSTATE Indication.

--brian


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