Tony Hain wrote:
>
> Alex Conta wrote:
>
> > I strongly believe, and am not the only one, please see
> > earlier messages
> > in the thread, packet forwarding and QoS processing in line
> > cards on routers at wire
> > speed is too critical to afford the use of flow label in the network
> > layer header, to pass duplicate information from source to
> > destination,
> > that has no network layer meaning, or use in the network.
>
> And there are probably more of us who strongly believe that there
> must be header bits that have immutable context from end-to-end.
> The network elements are not the drivers here, the endpoints are.
> If we allow the middle boxes to control our actions we end up with
> the POTS network, and loose the ability to deploy new and
> innovative applications.
>
> Tony
I agree, and you are right, we do not have to get to a place, where the
network controls everything.
In the same time, we have to be careful to not overload the main network
layer header with functions or information that are not critical to fast
header processing in the network.
I think with flexibility, and wisdom, we can accommodate any end-node
needs. IPv6 has this wonderful thing - we all know (-: - called
"destination options" headers that can carry information relevant to the
end-nodes, if that information is not pertinent to packet forwarding or
QoS in the network. There are also the transport and application
headers, which are just for end-node consumption.
For information that is pertinent, and critical to both network and end
nodes, for functions related to the network, we should make every effort
to include them in the main header.
Alex
S/MIME Cryptographic Signature
