RE: Ethernet Control Plane [Was: Re: Frameformat in a l2cs gmpls rnvironment.]

[<neil.2.harrison@bt.com>]

Tom,

G.805 (co) and G.809 (cl) discuss general functional architectural
properties of the data-plane network.  Specific technology ones are then
generated from these.  We've found this stuff invaluable for many
reasons....but its esp crucial if you want sound management information
models.....and for large operators esp this is a very important topic
IMO.   I find it most useful however in a day-day sense (at least in
terms of the principles its based on) to be able to quickly analyse a
technology and spot its strengths/weaknesses.  

Of course a data-plane network can carry traffic or control/management
protocols.  When you are dealing with co mode networks its quite
sensible (and forced in case of the co-cs mode) to have different
data-plane networks working in the same network system.  But see my post
to Igor a while back wrt to what this means when adapting these to be
carried by a lower layer network.  There are rather important network
owenship issues here.

Note also the design (eg for survivability) of the traffic and
control/management data-plane networks can be very different.  And it is
particularly important that the control/management-plane networks only
take their design cues from the duct-graph (you must know about real
physical separacy here if you want to do a good job)

The stuff in SG15 on ASON (various Recs here) discuss control-plane
arch.  And there is a stuff in the M series Recs on management
arch......but I'd look to TMF for this.

If you want to know a good text book, I've still to find anything better
than Reid/Sexton 'Broadband networking'.  Chapters 3 and 4 therein give
an excellent introduction to func arch and what this means for the 3
planes.....but there is lots of other good stuff in here on many topics.
Bit old-hat wrt technologies discussed, but this is not so importnat as
the general networking principles are invariant of technology (provided
one understands the basic modal differences).

regards, neil

> -----Original Message-----
> From: Tom Petch [mailto:nwnetworks@dial.pipex.com] 
> Sent: 23 July 2005 16:58
> To: Harrison,N,Neil,IKM1 R
> Cc: ccamp@ops.ietf.org
> Subject: Re: Ethernet Control Plane [Was: Re: Frameformat in 
> a l2cs gmpls rnvironment.]
> 
> 
> Neil
> 
> Is there an ITU-T Recommendation that gives an architectural 
> definition of what it means by control plane, data plane, 
> management plane, user plane, traffic dataplane, control data 
> plane etc etc?
> 
> I know what I mean when I use the terms but am not sure I 
> always know what others mean.
> 
> Tom Petch
> ----- Original Message -----
> From: <neil.2.harrison@bt.com>
> To: <John.E.Drake2@boeing.com>; <adrian@olddog.co.uk>; 
> <richard.spencer@bt.com>; <juergen.heiles@siemens.com>; 
> <loa@pi.se>; <per@defero.se>
> Cc: <ccamp@ops.ietf.org>
> Sent: Saturday, July 23, 2005 2:15 PM
> Subject: RE: Ethernet Control Plane [Was: Re: Frameformat in 
> a l2cs gmpls rnvironment.]
> 
> 
> > > > IP is cl-ps networking techology period....and its
> > > > jolly important, but PLEASE don't try and feed me any 
> of this 'IP 
> > > > control plane' nonsense.
> >
> > JD:  Over the years I have enjoyed Neil's rants immensely, 
> but I have 
> > to say that this one is a classic.
> 
> NH=> Thanks John for the kind words....its good to know 
> you've had at least something to smile about ;-)
> 
> Seriously though, see my updated response to Adrian.  But if 
> you are still struggling to grasp what I mean by the above 
> (out of context
> quote) try this:
> 
> A single cl-ps IP network carrying traffic and everything 
> else (eg public Internet) is very different animal to a co-cs 
> network system where the traffic data-plane network and 
> control/management data-planes network are logically 
> disjoint, ie the role of IP in the latter is importantly 
> different to the role of IP in the former...and let's hope 
> never the 2 shall meet.
> 
> regards, neil
> 
> 
>