[c-nsp] flow spec

[Roland Dobbins]

On Dec 3, 2006, at 4:37 PM, Yann Berthier wrote:

>    btw, what i've heard recently from another cisco guy was perfectly
>    aligned with what Roland Dobbins just stated in his reply - perhaps
>    there has been some shift in cisco's position recently regarding  
> flow
>    spec ?

Our position's been consistent since our participation in the initial  
flow-spec discussions in 2003 - we view flow-spec as a useful  
innovation in terms of basic control-plane-based signaling and  
mitigation functionality, which is why we have invested time and  
effort in co-authoring the draft.  This is still our view; we know  
from both SP and large enterprise customers that there are additional  
requirements above and beyond the objectives of the flow-spec draft  
(like other practical drafts, the flow-spec draft is focused on a  
very specific set of functionality, implemented in a fairly specific  
context), but we think flow-spec is a logical and operationally  
useful improvement on the basic S/RTBH concept (i.e., more  
granularity of filtering).

If flow-spec is important to you, please let your account-team and  
your internal purchasing group know this in very specific terms.  Its  
relative importance in relation to other features you require, the  
platforms/trains you want to see it on, its importance to you in  
evaluating new purchases, its inclusion as an RFP requirement, etc.  
are all very helpful to the folks who determine feature  
prioritization and scheduling.

-----------------------------------------------------------------------
Roland Dobbins <rdobbins at cisco.com> // 408.527.6376 voice

		All battles are perpetual.

     		   -- Milton Friedman