Re: Upper boundary of Adaptation Layers

Brian,
   
  I can only confirm what I told Barry!
   
  ITU-T specifications are very clear and I have read the ITU-T Q.1400 specification and there is nothing contradictory there to my point that each SCCP-user function controls its own traffic state -> there is no other way but to leave the user-function to control this since it is application dependent (i.e. each SCCP-user application can have its own management).
   
  ITU-T Q.1400 ch. 7 "Management Functionality":
  
"Figure 10 shows that each entity in the protocol architecture has a Level Management Entity (LME) associated with it.
The LME may be more or less explicitly defined within the level. For example, in the SCCP, there are a number of
management functions (e.g. SSN management) which constitute the LME for the SCCP."
   
  I do not have any problems with this since indeed SCCP management (like any other piece of functionality) has to be configured. For example SCCP Management must be configured in order to know to which remote SCCP subsystems the SCCP Management has to send SSP when it performs SSP broadcast. Another example is that "SCCP Signalling point congested" procedure as part of SCCP Management has also to be configured by the SCCP LM.
   
  However ITU-T Q.1400 certainly does not tell you that SCCP owns the functionality which is to remove the local SCCP-user form the service (or put it into service). Specifications are clear and in accordance to pure logic -> only SCCP-user entity can control logic which is to check if/how/when it can withdraw from traffic or start handling traffic!
   
  See ITU-T Q.711 "Functional description of the signalling connection control part", sec. 6.3 "SCCP management", sub-sec. 6.3.2.3.2 "STATE",
  
"The "N-STATE request" primitive (Table 16) is used to inform the SCCP management about the status of the originating user. The "N-STATE indication" primitive is used to inform an SCCP user accordingly."
   
  Note that N-xxx means Network Layer primitive (according to OSI model) i.e. primitive from Network Layer to its user function!
   
  / stanislav
   
  

"Brian F. G. Bidulock" <[email protected]> wrote:
  Stanislav,

Stanislav Ivanovich wrote: (Mon, 09 Jan 2006 09:32:04)
> 
> Barry,
> 
> 
> 
> I agree that user applications are to be isolated from the underlaying
> layers (i.e. SCTP, IP etc...). However I completely disagree that as
> you say both (!) entities (i.e. MTP/SCCP and their users) control the
> acitvation/deactivation of applications in the traffic! Instead in the
> legacy systems (and even in pure logic) it is in application
> responsiblity to know if/how/when to start/stop handling traffic.
> 

See Figure 10/Q.1400 and then reconsider Barry's answer.

--brian



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


		
---------------------------------
Yahoo! Photos
 Ring in the New Year with Photo Calendars. Add photos, events, holidays, whatever.

_______________________________________________
Sigtran mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/sigtran