[ NNSquad ] Re: Cisco Eyes Software Network Revolution

[Seth Johnson <seth.p.johnson@gmail.com>]

Geesh. No: the distinction between data and control is already in the
original TCP/IP protocol -- with control manifest in the form of a TOS
field nobody could really figure out what to do with (until the
*Inter*net became an IncumbentNet), and the TCP algorithms.

All that's new here is programming control algorithms for individual
routers.  There's no gain in openness here, when the control
algorithms they decide to do can only work across routers that agree
to do the same thing: that is, inside one network.  Let them try to
come up with something more than Internet2 was able to come up with,
but we can't hope for that so long as we still have an IncumbentNet.

It's misleading to call this open -- if that's taken to mean more open
than what we already have as far as the abstract protocols we
developed by general consensus.  This just lets network providers
experiment, not end users.  It's more for network providers to develop
specialized services than for providing the general purpose platform
for end users, across independent networks, that we got in 1977 when
they separated data from control by splitting the IP layer from TCP.
(Or, to consider recent revelations by Auerbach and Cerf, they came up
with a packet layer in secret NSA research in 1975.  This is not
mentioning Pouzin, etc.)


Seth Johnson


On Fri, Mar 16, 2012 at 7:49 PM, Dave Farber <dave@farber.net> wrote:

    ---------- Forwarded message ----------
    From: "Doug Humphrey" <doug@joss.com>
    Date: Mar 16, 2012 8:41 PM
    Subject: Re: [IP] Cisco Eyes Software Network Revolution
    To: <dave@farber.net>
    Cc: "ip" <ip@listbox.com>

    So the data world finally catches up to the phone guys - they did this
    long, long ago with CCIS (common channel interoffice signaling) and
    SS6 and SS7 (signaling system 7) - separating the data from the control.

    doug

    http://www.alcatel-lucent.com/bstj/vol57-1978/articles/bstj57-2-225.pdf

    http://en.wikipedia.org/wiki/Common_Channel_Interoffice_Signaling

    http://en.wikipedia.org/wiki/Signalling_System_No._7


    On Mar 16, 2012, at 6:26 PM, Dave Farber wrote:

    > ---------- Forwarded message ----------
    > From: "Lauren Weinstein" <lauren@vortex.com>
    > Date: Mar 16, 2012 6:20 PM
    > Subject: [ NNSquad ] Cisco Eyes Software Network Revolution
    > To: <nnsquad@nnsquad.org>
    >
    >
    > Cisco Eyes Software Network Revolution
    >
    > http://j.mp/A1r0Fn  (Wired)
    >
    >    "Essentially, OpenFlow separates networking into one plane that handles
    >     data and another that controls it movement. This is the way cellular
    >     networks have worked for years, but it was revolution in the data
    >     center networking business. The control plane could be run on standard
    >     servers, and then the data center plane could be run by fairly
    >     ordinary high-speed networking chips - rather lots of fancy new
    >     hardware."
    >
    >  - - -
    >
    > --Lauren--
    > NNSquad Moderator
    > _______________________________________________
    > nnsquad mailing list
    > http://lists.nnsquad.org/mailman/listinfo/nnsquad
    > Archives  | Modify Your Subscription | Unsubscribe Now
_______________________________________________
nnsquad mailing list
http://lists.nnsquad.org/mailman/listinfo/nnsquad