RE: [PWE3] Koleyni and Bocci drafts

[neil.2.harrison@bt.com]

Hi Khalid,
-----Original Message-----
From: Khalid Ahmad [mailto:kmad@nortelnetworks.com]
Sent: 21 June 2002 16:56
To: neil.2.harrison@bt.com; tdwalsh@lucent.com; harford@atmware.com;
pwe3@ietf.org
Cc: john.rutemiller@marconi.com
Subject: RE: [PWE3] Koleyni and Bocci drafts

<snipped>
I am still not so clear what means layer violation in this context, since in
my understanding the layer network modeling does not specifically address
interworked connections.

NH=> This is where a given technology (layer network) makes certain
assumptions/demands of either it's client layers or its server layers, and
thereore creates inter-layer linkages which hamper development/scope/usage
of technologies .....and in this context I mean wrt to data-plane
fault-detection/handling mecahnisms.  In Jeremy's orginal draft we had a
very obvious layer violation of asking L1 SDH/Sonet FDI to proxy for not
being able to handle this at the ATM layer in port mode....John/I are
discussing a way to handle this....not perfect but more clean.  Asking IP's
ICMP to proxy for missing MPLS fault handling is another layer violation.
The former is  a server layer violation and the latter is a client layer
violation.  That's the sort of issues I meant...it's simply bad
network/protocol architecture.  However, for most cases this area is simply
not adequately addressed in any of the drafts.    
regards, Neil
<snipped to end NH>

_______________________________________________
pwe3 mailing list
pwe3@ietf.org
https://www1.ietf.org/mailman/listinfo/pwe3