RE: ASP Active

Shashank,

I think from interoperability point of view, what ASPAC means is: "I am
ready to send/receive traffic". If a node sends an ASPAC when there is no
application ready to process the traffic, it won't break anything from M3UA
protocol point of view (M3UA-stack will receive messages and send them
upstream) but sending ASPAC in such a situation would be probably a bad
decision. So, I would exptect a M3UA-stack to be "SCCP-User aware" directly
or indirectly.

We may try to capture that point in the M3UA deployment considerations
draft -although it is more related with implementation rather than with
deployment-.

    Thanks,
    Tolga

> -----Original Message-----
> From: Prasad, Shashank S (Shashank) [mailto:[email protected]]
> Sent: Friday, December 23, 2005 9:11 AM
> To: 'Tolga Asveren'
> Cc: '[email protected]'
> Subject: RE: [Sigtran] ASP Active
>
>
> Tolga,
> I understand ur practical perspective.
>
> However, I am more concerned from the inter-operability
> perspective, between
> two IPSPs with different interpretations/implementations of
> when-to-send-ASPAC, I mean, one IPSP is "SCCP-User aware" while
> the other is
> NOT. The "SCCP-User aware" IPSP might expect the peer to also be
> "SCCP-User
> aware", which might not be the case.
>
> I believe that we have document a common interpretation on this,
> rather than
> leaving it open ended.
>
> Thanx.
> shashank
>
> -----Original Message-----
> From: [email protected] [mailto:[email protected]]On
> Behalf Of Tolga Asveren
> Sent: Friday, December 23, 2005 6:43 PM
> To: [email protected]
> Subject: RE: [Sigtran] ASP Active
>
>
> Shashank,
>
> IMO, when a node initiates ASPAC is more related with internal
> implementation -also the relationship of LM with other entities
> and how that
> functionality is implemented-, from M3UA-stack point of view it just means
> that the node is ready to send/receive traffic for the coreresponding
> traffic range and M3UA-stack should allow sich traffic to pass in both
> directions, but practically without an application ready to process the
> messages this wouldn't have much meaning.
>
>      Thanks,
>      Tolga
>
> > -----Original Message-----
> > From: [email protected] [mailto:[email protected]]On
> > Behalf Of Prasad, Shashank S (Shashank)
> > Sent: Friday, December 23, 2005 6:59 AM
> > To: 'Ilie Glib'
> > Cc: SIGTRAN; [email protected]
> > Subject: RE: [Sigtran] ASP Active
> >
> >
> > Ilie,
> > I was trying to answer/question Tolga's and Stanislav's point, when they
> > said that ASPAC declares readines of the application, i.e.
> SCCP-User in my
> > example.
> >
> > I am of the opinion that M3UA should NOT be check the "readiness" of
> > SCCP-User, before sending the ASP Active. Or for that matter, it
> > should NOT
> > even be aware of the states of SCCP-User
> >
> >
> > shashank
> >
> >
> >
> >
> > -----Original Message-----
> > From: Ilie Glib [mailto:[email protected]]
> > Sent: Friday, December 23, 2005 4:56 PM
> > To: Prasad, Shashank S (Shashank)
> > Cc: Stanislav Ivanovich; [email protected]; SIGTRAN
> > Subject: Re: [Sigtran] ASP Active
> >
> >
> > Hello Shashank,
> >
> > M3UA should not have any knowledge of SCCP-Users, perhaps you meant
> > something different in your statement below. M3UA knows whether SCCP
> > is up and running, that is made visible by SG via DUPU, but it is not
> > necessary to be visible on IPSP to IPSP interface (M3UA), since SI is
> > not mandatory part of the RK, and RC in general is optional. Handling
> > of SCCP-Users is up to SCCP layer, in your network scenario.
> >
> > Ilie
> >
> > On 12/23/05, Prasad, Shashank S (Shashank) <[email protected]> wrote:
> > > Hi Stanislav, Tolga,
> > >
> > > Well, then that would mean that Layer Manager (LM) needs to
> > keep track (or
> > > be aware) of SCCP and applications over SCCP to initiate the
> ASP Active
> > ie.
> > > LayerN needs to be "aware" of the states of Layer N+2.
> > >
> > > The applications (SCCP-User) should have inherent ways to
> > inform the peer
> > if
> > > it's NOT ready, OR to drop the incoming message.
> > > And the peer -to-peer messaging between SCCP-Users, should
> take care of
> > > invalid states of one another.
> > >
> > > shashank
> >
> > _______________________________________________
> > Sigtran mailing list
> > [email protected]
> > https://www1.ietf.org/mailman/listinfo/sigtran
> >
>
>
>
> _______________________________________________
> Sigtran mailing list
> [email protected]
> https://www1.ietf.org/mailman/listinfo/sigtran
>