Re: [Idr] Genart last call review of draft-ietf-idr-rfc5575bis-20

[Christoph Loibl <[e-mail]>]

Hi Gyan,

Thanks for your review. According to your review I made the following changes to the document which is available now as revision -22:

> On 07.04.2020, at 23:43, Gyan Mishra via Datatracker <[e-mail]> wrote:
> 
> Reviewer: Gyan Mishra
> Review result: Ready with Nits
> 
> Reviewer: Gyan Mishra
> Review result: Ready with Minor Issues
> 
> Minor issues:
> I am familiar with BGP Flow specification and would like to recommend some
> verbiage that may help in the introduction as far as explaining how BGP flow
> spec works.  Ssince the introduction has been re-written with this update, this
> could be a possible addition to the draft.
> 
> This could be placed at the end of the introduction if desired.
> BGP flow specification is a client-server model that allows for a more granular
> approach to DDOS mitigation than its predecessor, “Remotely Triggered Blackhole
> (RTBF) which tagged a prefix with a community and sent it do a discard next
> hop.  BGP flow spec has two main components, the “controller” being the BGP
> speaker device which acts as the server side, which injects the new flowspec
> entry, and the client side which is the BGP speaker devices that receives the
> flowspec NLRI and acts on the instruction to match a particular flow with Layer
> 3 and Layer 4 parameters and then implements the hardware forwarding action
> requested.

<-- 
Tracked via issue #163: https://github.com/stoffi92/rfc5575bis/issues/163

I do not agree that BGP flowspec is a client-server model -only-. We can propagate this NLRI over administrative domain borders as we do with IP routing information, it follows the same mechanisms. We see such solutions being deployed in the internet as inter provider DDoS solutions.

We actually had a paragraph in the darft that was explaining the advantages over other approaches like RTBF but this has been removed because it was pointed out that it is not relevant to the spec to justify a well deployed technology.
-->


> Nits/editorial comments:
> 7.  Traffic Filtering Actions
>   This document defines a minimum set of Traffic Filtering Actions that
>   it standardizes as BGP extended community values [RFC4360]
> 
>   Any mention of [RFC4360] should be updated with [RFC7153] IANA Registries
>   for BGP Extended Communities.
> 

<-- 
Tracked via issue #164: https://github.com/stoffi92/rfc5575bis/issues/164
Commits mentions:
    https://github.com/stoffi92/rfc5575bis/commit/31f0ac79b7cd998aa2750fd376dc148d7a590369
    https://github.com/stoffi92/rfc5575bis/commit/7aadadcdf55a1f5a7d5c1822070b862247dfaead

Removed the "values" statement (as suggested by Alvaro) from the draft to make clear we are not talking about particular values but about  Extended Communities as specified in RFC4360.
s/standardizes as BGP extended community values [RFC4360]/standardizes as BGP extended communities [RFC4360]/

-->

Cheers 
Christoph

-- 
Christoph Loibl
[e-mail] | CL8-RIPE | PGP-Key-ID: 0x4B2C0055 | http://www.nextlayer.at