Re: [lmap] Feedback on draft-eardley-lmap-terminology

[Juergen Schoenwaelder <[e-mail]>]

On Wed, Jul 24, 2013 at 01:43:53PM +0000, Romascanu, Dan (Dan) wrote:
> 
> One of the points where the framework is still confusing for me and the terminology I-D does not help is the controller.
> 
> Says the terminology I-D:
> 
> Controller: A function that provides a Measurement Agent with
>    Instruction(s).  Colloquially, a Controller is a physical device that
>    performs this function.
> 
> Do we really envision physical devices that perform this (only) function?

For me, the second sentence is rather pointless. I see not reason why
a controller may not be a virtual machine. The controller is a
function, it does not really matter how it is realized. Note that we
have similar sentences for the Collector and Measurement Agents (and I
think they all three could go).
 
> More unclarity. The single controller (for a given MA) key assumption is not mentioned by draft-akhter-lmap-framework. Why? The other framework draft draft-eardley-lmap-framework doest take into account the key assumtion, but then talks about the MA pulling configuration from the Controller in order to avoid firewall traversal problems. Does this eliminate a-priori the usage of NETCONF as a possible transport for the Control Protocol?
> 

Perhaps things need to be rephrased but then I am not sure which
sentence in draft-eardley-lmap-terminology-02 you think is ruling out
NETCONF. The underlying requirement is that sessions need to be
initiated by the MAs since they may be behind NATs or other
middleboxes.

/js

-- 
Juergen Schoenwaelder           Jacobs University Bremen gGmbH
Phone: [phone]         Campus Ring 1, 28759 Bremen, Germany
Fax:   [phone]         <http://www.jacobs-university.de/>