[Paper] B4: Experience with a Globally-Deployed Software Defined

Jayram Déshpandé jaydesh9 at gmail.com
Sun Aug 18 00:26:07 UTC 2013


SDN is not a new concept at all.

Infact since ARPANET days, the notion of centralized control plane had a
lot of traction. But with Cold war around, It made more sense to push the
control plane intelligence into individual decision points (routers ,
switches , et . al. ). Considering the possibility of the commies taking
down some part of the early Internet, the remaining partitioned network
could still survive as the rest of the decision points could converge and
act as independent network snippets.

-Jay.



On Sat, Aug 17, 2013 at 4:29 PM, Jeff Kell <jeff-kell at utc.edu> wrote:

> On 8/17/2013 7:14 PM, Arturo Servin wrote:
> >       Hacker will love SDN ...
>
> Yes.  Traditional SDN is big, flat layer-2 network with global
> mac-address resolution, and a big fat Java applet managing the adjacency
> tables.
>
> What could *possibly* go wrong?
>
> Jeff
>
>
>


-- 
"Subvert the paradigm." - C.K. Prahlad


More information about the NANOG mailing list