Re: Disable Exponential Backoff (retry) on Ethernet?

If I have to, I will employ some form of simulation of collisions, but that
will have to be done carefully in order to get the effect of collsions
beteen multiple stations without messing up timing.

I REALLY wanted to have the PHY handle it. Usng half duplex and some old
hubs (if I can find one) would get collisions, but the retry would create
other problems.

So, looks like "simulated" collisions will be the way to go.

BTW: I'm using Netgraph - very cool stuff.

--Len




On 10/20/07, Harti Brandt <hartmut.brandt@xxxxxx> wrote:

On Sat, 20 Oct 2007, Peter Jeremy wrote:

PJ>On Wed, Oct 17, 2007 at 08:51:34PM -0700, Len Gross wrote:
PJ>>I'm doing some protocol development and it is convenient to start it
on
PJ>>Ethernet. I will need to send a packet to the Ethernet device and
only have
PJ>>it be sent once, even if there is a colision.
PJ>
PJ>I know we've still got some hubs lying around in a backroom at work
PJ>but I don't know of anything that will let you disable the retry-on-CD.
PJ>
PJ>Have you considered simulating the network at a slightly higher lever:
PJ>Use ipfw pipes or similar to simulate packet loss (either set a queue
PJ>length of 1 or probabilistically). This could be done either as a
PJ>bridge or by tunneling your protocol over IP or UDP.

Some years ago I wrote a netgraph node that connected to ethernet nodes
and simulated a wireless broadcast channel including collisions,
timevariable delay, loss and shaping. Can be done in a couple of hundreds
of lines and easily allows >100MBit/sec with gigabit ethernet. The
ethernet is just the physical transport medium for the packets and does
not take part in the emulation. All that was controlled by bsnmpd and a
remote command line tool and/or Java-GUI. Unfortunately I cannot release
this (yet) due to licensing...

harti

_______________________________________________
freebsd-net@xxxxxxxxxxx mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-net
To unsubscribe, send any mail to "freebsd-net-unsubscribe@xxxxxxxxxxx"

Relevant Pages

  • Re: Disable Exponential Backoff (retry) on Ethernet?
    ... looks like "simulated" collisions will be the way to go. ... PJ>Have you considered simulating the network at a slightly higher lever: ... PJ>Use ipfw pipes or similar to simulate packet loss (either set a queue ... Some years ago I wrote a netgraph node that connected to ethernet nodes ...
    (freebsd-net)
  • Re: Ethernet under high load
    ... Efficient and accurate ethernet simulation. ... If a node has been unsuccessful at sending a frame, it backs off a little longer for each failed attempt, until the 10th attempt, then use the long max wait of the 10th attempt for the remaining 5 attempts. ... The closer you get to 1024 nodes per Ethernet, the more likely it will be for nodes to have the same retry wait, and to experience collisions. ...
    (comp.dcom.lans.ethernet)
  • Re: Collision on NIC
    ... >> Even Mr. Inventor of the ethernet himself regrets calling them ... If it were a problem there might be a solution, but collisions ... they're normal on an ethernet. ...
    (freebsd-net)
  • Re: Disable Exponential Backoff (retry) on Ethernet?
    ... type protocol) so I actually need collisions! ... I will need to send a packet to the Ethernet device and only ... Are you even using hardware that does collisions? ... 10/100mbit "switches" that only switched between the two segments (and ...
    (freebsd-net)
  • Re: Ethernet under high load
    ... Efficient and accurate ethernet simulation. ... where it simulates the network characteristics accurately, the offered load ...
    (comp.dcom.lans.ethernet)