Re: Control plane resiliency

["Igor Bryskin" <ibryskin@movaz.com>]

George,



Although I do agree that mechanisms to control IP and non-IP networks should
not fundamentally deviate, we still need to bear in mind that there are
things that work well for IP networks and not so well or sufficiently well
for non-IP networks and visa versa. For example, the control and data planes
are congruent in IP and, generally speaking, not-congruent in transport
networks, which brings additional challenges  wrt resiliency/robustness of
the control plane that we have never met in the past.



I recommend to read the paper Diego mentioned in his email, where the
authors described quite accurately some of the problems that I have referred
to in the thread.



Regards and see you on IETF64.

Igor



----- Original Message ----- 
From: "George Swallow" <swallow@cisco.com>
To: "Adrian Farrel" <adrian@olddog.co.uk>
Cc: <ccamp@ops.ietf.org>; <swallow@cisco.com>
Sent: Monday, October 31, 2005 6:42 PM
Subject: Re: Control plane resiliency


> Adrian -
>
> Excellent note.  In reading through the thread I was beginning to fear
> that having chosen IP as a means of providing a resilient control plane,
> we might totally complicate things and fundamentally deviate from IP at
> this late date.
>
> I believe the exercise of working through requirements, showing how
> existing mechanism can and SHOULD be used to address these, and in the
> process point out any omissions would be a very useful piece of work.
>
> Thanks,
> ///George
>
> ========================================================================
> George Swallow             Cisco Systems                  (978) 936-1398
>                            1414 Massachusetts Avenue
>                            Boxborough, MA 01719
>