The IP over ATM Mailing List Archive by date

[murray@pa.dec.com]

> but i guess since it only happens at a merge point for mp-p
> conenctions, it only ever introduces 1 of these frame
> store-and-forward serialisation times, rather than n for n hops...

Nope.  You add a store and forward delay at each merge point.  For 
a big tree, there could be several/many of them.


> ... whilst the latter removes the cut
> through store and forward  advantage of forwardign a cell at a
> time

You don't have to take the store-and-forward delay all the time, 
just when there is another packet already using the outgoing VC.

You can forward the first packet cell at a time.  If a second packet 
arrives, you buffer it until the first packet ends.