The IP over ATM Mailing List Archive by date

[Masataka Ohta <mohta@necom830.cc.titech.ac.jp>]

> > > 1. Aim for good layer 2 redundancy/failover (e.g. use PNNI signalling)
> > >      and use a simple upper layer protocol (e.g. RFC 1577)
> > 
> > It can't be redundant against the ARP server failure, the
> > single-and-only link failure to the ARP server or the single switch
> > failure to which the ARP server is directly attached.
> 
> With PNNI and Anycast signaling to provide failover to alternative servers
> and with a server-server coordination protocol, a very simple upper-layer 
> protocol e.g. RFC 1577 between client and server is sufficient.

I'm afraid you are suggesting to have RFC 1577, server-server
coordination protocol, broadcast emulation protocols, multicast
emulation protocols and so on none of which are proven to work.

> > As LANE will be there, why do you think we need work to have RFC826-?
> 
> As I've stated here before, that could be the $10000000 question ....

What ever the cost of LANE based solutions, which use all the existing
and proven-to-work protocols, it is at least 10 times LESS than that
of RFC 1577 based ones, I think.

And, don't forget that,

> 2. Assume VCCs can come and go with poor failover mechanisms (e.g. UNI 3.x) 
>      and define a more robust upper layer protocol (e.g. LAN Emulation)

rubust upper layer protocols can be even more rubust with PNNI.

							Masataka Ohta