Im looking at accurate network models made by people who have spent
their life doing it, and whose models fairly accurately predict current
network patterns
> 1. It's not all TCP/IP. There are many protocols without backoff.
> If the TCP/IP connections do the backoff, other protocols will
> have an advantage and will consume all bandwidth. TCP/IP won't
> get a fair share by being polite.
Their volume is microscopic, and their behaviour is generally backoff
> 2. The exponential backoff can not be enforced, since users can
> restart the connection. That is exactly what happens.
Their restart also resets the congestion window to one frame
> 3. The last time I heard, http was the #1 protocol. Since the
> connections are short-lived, backoff works poorly anyway.
http/1.1 uses persistant connections
Alan
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/faq.html