Re: Active networking in ns

[Polly Huang <huang@catarina.usc.edu>]

On Wed, 13 Oct 1999, Guillermo Rodriguez Garcia wrote:

> Hello,
> 
> I am developing a new protocol to be used in active
> networks (for an introduction to active networks,
> see http://www.tns.lcs.mit.edu/activeware). The main
> issue in active networks is that "active nodes" in
> the path from the source to the destination might
> do some (user-specified) additional processing on
> the protocol packets.
> 
> I have written a new agent and I attach it to the
> nodes I want to be active, but the problem is that
> I am not sure on how to allow intermediate nodes
> to process to packet.
> 
> For example, let's say I have the following simple
> topology:
> 
> n0(src) ---- n1 ---- n2(dst)
> 
> I attach active agents to all the nodes (n0, n1, n2)
> and then send a packet from n0 to n2. The problem is
> that the agent at n1 won't see the packet. I guess
> that I could modify classifier.cc so that my packets
> are always evaluated at each node, but I would like
> to know if there is any way to do this without
> modifying ns itself too much. Any suggestion ?

Guillermo,

DOes n1 intercept the packet, process it, and then decide what to
do next (for example forward it down to n2) or n1 make a quick copy of the
packet and process it, while the packet continues to be fowarded down
to n2? If it's the latter, you can probably use a multicast address for
your active network agents in n1 and n2 both to join.  That way, both
agents are sure to receive the packet.  This is a quick and not very
pretty hack, but you can get away with digging into the classifer.cc
details.

just a thought,
-Polly

> 
> Regards,
> G.
> -- 
> Guillermo Rodriguez Garcia
> <guille@iies.es>
>