RE: RFC 4666 M3UA - Reg Request Message

Brian,

If you look at LOADSEL from ONLY CIC perspective, all I see that a new parameter is defined "CIC RANGE" which can very well be defined within Routing Key defined in 4666. 

Again talking ONLY from CIC perspective, I don't know as what problems you are trying to quote which are solved by LOADSEL and not by re-introducing CICs at Reg-Req.

I agree to the fact that LOADSEL takes a generic approach for Load sharing but I disagree load sharing when comes to a very CIC level. Each CIC is a unique physical resource between any OPC/DPC. All call handling messages for any OPC-DPC-CIC resource should be handled by a unique Active ASP.

Again notification to other ASPs regarding resources which are never handled by them is the same as such resources are given a generic treatment at SG (or by Default ASP as RFC says). I don't understand as why this is a better solution? 
Relocating ISUP handling to SG or default ASP is implementation dependant if you are trying to avoid that by this method.

I also don't understand as what other three problems pertaining to CIC & ISUP traffic is solved by LOADSEL? 

LOADSEL is a good draft from various load sharing parameters at various protocols but I still think it does not solve any issues for ISUP and traffic at CIC Level.

-Jitin

-----Original Message-----
From: Brian F. G. Bidulock [mailto:[email protected]] 
Sent: Tuesday, November 07, 2006 2:35 PM
To: Bhandari, Jitin (Jitin)
Cc: 'Ilie Glib'; Haresign Lincoln; [email protected]
Subject: Re: [Sigtran] RFC 4666 M3UA - Reg Request Message

Jitin,

Bhandari, Jitin (Jitin) wrote:                     (Tue, 07 Nov 2006 14:25:57)
> Thanks Brian for the responses. I did read the summary section and it is
> just that I was trying to confirm couple of thoughts being exchanged in this
> email chain.
> 
> Anyhow, I agree with Ilie that either by re-introducing CIC Ranges for
> Reg-Req or defining LOADSEL draft as an extension could be a viable path.
> 
> However, I still think that LOADSEL is not solving any issues of un-handled
> CICs when AS is unavailable so if CICs are re-introduced in Reg-Req message
> for ISUP traffic, re-introducing CIC ranges in Reg-Req is the simplest way
> to solve handling of ISUP traffic. 

I disagree.  Routing on CIC introduces more problems, LOADSEL less.

> Having a default ASP handling un-available CICs is same as having an
> implementation dependant solution at SG. I think Notify procedures between
> ASPs also don't solve this issue.

It is a far better situation with LOADSEL where ASPs are informed that a
CIC range is unserved.

> Talking about requirements, If CIC handling (Note ONLY CIC Handling between
> ASPs) is the only issue to be solved, then for a simplest solution I see
> re-introduction of CIC ranges is the best way. Though for now I see CIC
> handling amongst ASP as a bigger issue given provisioning nature of CICs at
> MGC.

It is certainly not the simplest approach and it does not address the three
other issues addressed by LOADSEL.

--brian


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