Re: [Idr] AD Review of draft-ietf-idr-bgp-extended-messages-20

["Susan Hares" <[e-mail]>]

Alvaro:

<shepherd hat on> 
I'm saying your suggested revisions did not correctly align with RFC4271 specification, specifically in the area of the FSM.  I objected to version-21 because it contained errors by misaligning the text with the IDR FSM.  (see mail list for specific issues0.  I proposed a cleaner replacement text.  Earlier versions did not hit this misalignment.  As a shepherd, I felt version-21 could not go forward.  
<shepherd hat off> 
<individual contributor, co-author on RFC4271 hat on> 
I am saying the better way is to just utilize the BGP message header length error function (0-4096 normal, 4097-65535 extended) and let the FSM handle it.  Then all BGP messages could be large. See the email thread. 
<individual contributor hat off> 
<WG chair hat on> 
This is why we have the thread on the mail list about accepting all messages
</WG chair hat off> 

Happy changing hat day to you! 
Sue
-----Original Message-----
From: Alvaro Retana (aretana) [mailto:[e-mail]] 
Sent: Tuesday, March 7, 2017 5:09 PM
To: Susan Hares; 'Randy Bush'
Cc: [e-mail]; [e-mail]; [e-mail]
Subject: Re: AD Review of draft-ietf-idr-bgp-extended-messages-20

On 3/5/17, 3:17 PM, "Susan Hares" <[e-mail]> wrote:

Sue:

Hi!

> A few point have raised my disagreement regarding the AD reviews.   At this
> point, the text is at odds with RFC4271.   If you feel RFC4271, does not
> speak about these issues - you may have missed why the BGP FSM is there.
> Please change your text. 

I’m not sure why you disagree with my comments…but if I interpret your points correctly, I think you’re saying that this document should also update the FSM where needed, is that right?  If so, then I agree with that.

Thanks!

Alvaro.