Re: [lisp] AD Review of draft-ietf-lisp-pubsub-09

[Dino Farinacci <[e-mail]>]

> On Nov 6, 2022, at 8:49 PM, Alberto Rodriguez-Natal (natal) <[e-mail]> wrote:
> 
> 
> Hi Dino,
>  
> I’m referring to this exchange [1] between Alvaro and you, regarding the use of ITR-RLOCs from the Map-Request as the addresses to use to send the Map-Notifies to.
>  
> The summary is that since Map-Requests include the “ITR-RLOC” fields, we are using those, instead of the source address of the Map-Request, to send the Map-Notifies. This is for both Map-Notifies used to ack subscription as well as for Map-Notifies sent as publication (the Map-Server keeps track of the last ITR-RLOCs received for a subscriber).
>  
> Hope it makes sense :)

It makes sense now but we have to be careful those ITR-RLOC addresses are translated public addresses so the Map-Notify can get back to ITRs through NATs. 

Dino

>  
> Alberto
>  
> [1] https://mailarchive.ietf.org/arch/msg/lisp/hN4tpuZl2nm86UVTqqmH24q0Tx8/
>  
>  
> From: Dino Farinacci <[e-mail]>
> Date: Sunday, November6 2022 at 2:39 PM
> To: Alberto Rodriguez-Natal (natal) <[e-mail]>
> Cc: Alvaro Retana <[e-mail]>, [e-mail] <[e-mail]>, Luigi Iannone <[e-mail]>, [e-mail] <[e-mail]>, Vina Ermagan <[e-mail]>, Stefano Secci <[e-mail]>, Fabio Maino (fmaino) <[e-mail]>, Sharon Barkai <[e-mail]>, Albert Cabellos <[e-mail]>, Johnson Leong <[e-mail]>, JACQUENET Christian INNOV/NET <[e-mail]>, [e-mail] <[e-mail]>
> Subject: Re: AD Review of draft-ietf-lisp-pubsub-09
> 
> > [AR] As mentioned in the Alvaro/Dino discussion, regular Map-Notifies as per 9301 are sent to the source IP address of the Map-Register that triggered them. Since PubSub extends LISP to specify that Map-Notifies can also be triggered by Map-Requests, we need to specify where these Map-Request-triggered Map-Notifies are sent to.
> 
> This isn't making sense to me. Are you talking about a Map-Notify as an ack to the Map-Register or an unsolicited Map-Notify and if the latter, the notification is going to the source address of the LAST subscription request. And if another Map-Request came later with the N-bit set again, with a new/different source IP address, the map-server must cache the last address (and later send Map-Notify to this new address).
> 
> Agree?
> 
> Dino