Re: [ncrg] Meeting Notes from Today's NCRG Call

On May 14, 2013, at 12:44 PM, "Alvaro Retana (aretana)" <aretana@cisco.com> wrote:

> On 5/14/13 11:29 AM, "Michael Behringer (mbehring)" <mbehring@cisco.com>
> wrote:
> 
> Hi!
> 
>> Attendees: Sheng Jiang, Xin Sun, Rana Sircar, Vladimir Marbukh, Michael
>> Behringer (notes)
> 
> Sorry I couldn't make the call..too many meetings. :-(

Same here...

> 
>> Subject: Evolving the NCRG framework draft
>> 
>> Brief notes: 
>> -	Suggestion to include metrics and tradeoffs in the framework draft;
>> this was presented in Paris.

If the framework includes all the forms, then including all tradeoffs and metrics of all forms in a single document will be a bit challenging, I suspect. The multitude of details will also deter from a simple framework.

>> -	draft-retana suggest more metrics, and trade-offs, these should also be
>> in the framework draft.
>> ..
>> -	Include text to describe the various forms (?) of complexity
>> 	+ Operational complexity
>> 	   deployment complexity; configuration complexity, trouble shooting
>> 	+ Network system complexity
>> 	  s/w complexity; h/w complexity; Protocol / algortihm complexity
>> 	+ Management system complexity
>> 	  OSS systems, etc.
> 

It seems there might be some overlaps and gaps on these various forms. 

> 
> Our draft (draft-retana-network-complexity-framework) shows tradeoffs
> (which may result in potential metrics) related to network design.
> 
> I agree that the framework should describe different types of
> complexity..network design complexity being one of them.  My only concern
> with including all forms of complexity, their description, metrics,
> tradeoffs, etc. in one document is simply scope creep: we'll try to cover
> too much and never finish.  The topic is already complex enough (pun
> intended)!
> 
> IMHO, I would prefer the main framework document to briefly describe the
> different forms (operational, design, etc.)..but then to have standalone
> documents that supplement it by going deeper into a specific form and
> describing the tradeoffs and maybe potential metrics.
> 

I agree with this proposal. I believe that the framework should provide the taxonomy, relationships, and definitions. But not the details and metrics. Companion documents could cover that.

Thanks,

-- Carlos.

> My 2c.
> 
> Alvaro.
> 
> _______________________________________________
> ncrg mailing list
> ncrg@irtf.org
> https://www.irtf.org/mailman/listinfo/ncrg
> 