Re: [NSIS] state management and the nsis framework

Hi John

What I am actually saying is that  both layers should be able to provide
soft state
management. The soft state management at both layers could use refresh
messages
belonging to one of the protocol levels, e.g., NTLP. If a NTLP state is not
refreshed
in time the NSLP instance is informed about it.
Otherwise the NSLP instance assumes that the NTLP state is alive.
This is similar to the concept proposed by Bob Braden,
see:
http://www.ietf.org/internet-drafts/draft-braden-2level-signal-arch-01.txt

In this way the the number of refresh messages
that have to be processed, per signaling application,
by the NTLP/NSLP aware node is minimized.

Best regards,
Georgios

----- Original Message -----
From: <john.loughney@nokia.com>
To: <karagian@cs.utwente.nl>; <luu@enst.fr>; <nsis@ietf.org>
Sent: Wednesday, May 14, 2003 12:19 PM
Subject: RE: [NSIS] state management and the nsis framework

> Hi Georgios,
>
> > When we only use refresh message belonging to one of the protocol levels
then
> > for one signaling application the number of refresh messages needed
> > can be two times smaller than the number of refresh messages that have
to be
> > processed when we use NTLP refresh messages to refresh the NTLP states
> > and independent NSLP refresh messages to refresh the NSLP states.
>
> At present, I don't know if we have made a determination which layer
> must refresh what layer. One could imagine a light-weight NTLP which
> does not require refreshign, but all is handled via NSLP refreshes
> or visa-versa.
>
> It seems that you are requesting the possibility at refreshing at
> both layers, correct?
>
> John
>

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis