Re: [Idr] I-D Action: draft-ietf-idr-as-migration-04.txt

Hi,

(on PTO with limitted internet access)

Possible updates:

- As we discussed, for ebgp incomming updates for local-as it makes more 
sense to prepend AS on incomming updates. I think we shoudl describe this 
option as the default one and any other one as a variation.

- Cisco plans to implement the internal BGP alias as ibgp local-as 
dual-as. So perhaps we can choose an agnostic name and explain both 
Juniper and Cisco names.

- About "the speaker SHOULD send BGP OPEN using the globally configured 
ASN first". I think we should emphasize how the FSM gets modified. It may 
be implicit, but it may save us from some interoperabilities problems. My 
ideas:

  1) In any case, my-AS and your-AS MUST be equal on any case.
  2) Define what to do if you send global-AS but you receive local-AS. 
Probably respond wiht "Bad Peer AS". But I'd like to understand what 
Juniper is doing as well.
  3) Note that without an agreement of the AS that is used first (both 
peers could decide to use first the local-as if they are both configured as 
such). This would happen in 2)

-Juan

On Mon, 20 Apr 2015, UTTARO, JAMES wrote:

> What would be great is that the features that are being used are standardized across vendors. The current approach where each implementation varies in behavior will make life difficult.
>
> Jim Uttaro
>
> -----Original Message-----
> From: Idr [mailto:idr-bounces@ietf.org] On Behalf Of Susan Hares
> Sent: Monday, April 20, 2015 3:53 AM
> To: 'Juan Alcaide'; 'Alvaro Retana (aretana)'
> Cc: idr@ietf.org; jgralak@juniper.net
> Subject: Re: [Idr] I-D Action: draft-ietf-idr-as-migration-04.txt
>
> Juan:
>
> Do you think something needs to change in this draft?  This is the last day
> of the WG adoiption call.
>
> Sue
>
> -----Original Message-----
> From: Idr [mailto:idr-bounces@ietf.org] On Behalf Of Juan Alcaide
> Sent: Thursday, April 16, 2015 5:27 PM
> To: Alvaro Retana (aretana)
> Cc: idr@ietf.org; jgralak@juniper.net; Susan Hares
> Subject: Re: [Idr] I-D Action: draft-ietf-idr-as-migration-04.txt
>
> Hi,
>
> Let me chip using a previous discussion inputs.
>
> When using local-as, Cisco prepends AS on inbound updates from eBGP peers.
> I understand Juniper prepends AS on outbounds updates towards eBGP peers.
> In most cases, Cisco approach is better.
>
>>>>> Consider
>>>>>
>>>>> R1 (AS1) ----(local-as 2) R2(AS3)--- R3(AS3) --- R4 (AS4)
>>>>>          ebgp                    iBGP
>>>>>
>>>>> Cisco:
>>>>>
>>>>> R1: X
>>>>> R2: 2 X
>>>>> R3: 2 X
>>>>> R4: 3 2 X
>>>>>
>>>>> Juniper
>>>>>
>>>>> R1: X
>>>>> R2: X
>>>>> R3: 2 X
>>>>> R4: 3 2 X
>>>>>
>>>>> I think it's prefereable that R2 and R3 have the same AS-PATH.
> Otherwise
>>>>> you can have inconsistent best-path decission inside AS3
>
>
>
> -Juan
>
> On Thu, 9 Apr 2015, Alvaro Retana (aretana) wrote:
>
>> On 4/9/15, 12:43 PM, "Susan Hares" <shares@ndzh.com> wrote:
>>
>>> Let me review the document and query Alvaro to make sure it addresses
>>> his concerns.
>>>
>>> On re-doing WG LC, we can do this in parallel with the other reviews
>>> I'll need to redo (security directorate, routing directorate,
>>> ops-directorate and gen-art). I'll start both sequences at the same
>>> time once I've heard from Alvaro.
>>
>> I¹ll go read it, but my gut reaction is that we should go back through
>> (WGLC, etc.).
>>
>> Thanks!
>>
>> Alvaro.
>>
>> _______________________________________________
>> Idr mailing list
>> Idr@ietf.org
>> https://www.ietf.org/mailman/listinfo/idr
>>
>
> _______________________________________________
> Idr mailing list
> Idr@ietf.org
> https://www.ietf.org/mailman/listinfo/idr
>