Re: [Idr] I-D Action: draft-ietf-idr-bgp-model-04.txt

["Rajiv Asati (rajiva)" <[e-mail]>]

Indeed.

It would be useful to have routing protocol independent structure for address-family and minimize the duplication in abstraction (and implementations).

--
Cheers,
Rajiv

From: Idr <[e-mail]> on behalf of "Acee Lindem (acee)" <[e-mail]>
Date: Wednesday, March 20, 2019 at 10:57 AM
To: Xufeng Liu <[e-mail]>, Mahesh Jethanandani <[e-mail]>
Cc: "[e-mail]" <[e-mail]>
Subject: Re: [Idr] I-D Action: draft-ietf-idr-bgp-model-04.txt

I agree – the BGP hierarchy is wrong. Address-family is generally higher both abstractly and in implementations.
Thanks,
Acee

From: Idr <[e-mail]> on behalf of Xufeng Liu <[e-mail]>
Date: Wednesday, March 20, 2019 at 10:30 AM
To: Mahesh Jethanandani <[e-mail]>
Cc: IDR List <[e-mail]>
Subject: Re: [Idr] I-D Action: draft-ietf-idr-bgp-model-04.txt

Hi Mahesh,

On Tue, Mar 19, 2019 at 6:48 PM Mahesh Jethanandani <[e-mail]<mailto:[e-mail]>> wrote:
Hi Xufeng,



On Mar 19, 2019, at 6:31 AM, Xufeng Liu <[e-mail]<mailto:[e-mail]>> wrote:

Hi Mahesh,

Thanks for the update.

I'd like to comment on the new changes at a high level:


1) rib extension

This paradigm is inconsistent with other protocol models like ospf and isis, where the protocol specific routes are kept under the protocol instance tree, not under the /rt:routing/rt:ribs. Based on RFC8349, the /rt:routing/rt:ribs tree is used to model the routes per routing instance, which is better mapped to the Route Manager (whose name varies depending on the implementations).


While that might be true, routes in the BGP model currently are maintained at the per-address family level.

It is fine that routes are maintained at per-address family level, which is also done by other routing protocols. The question is how the tree hierarchy is structured.
OSPF model has the following:

module: ietf-routing
  +--rw routing
  |  +--rw control-plane-protocols
  |  |  +--rw control-plane-protocol* [type name]
  |  |     +--rw ospf:ospf
  |  |        +--ro ospf:protected-routes {fast-reroute}?
  |  |        |  +--ro ospf:af-stats* [af prefix alternate]
  |  |        |     +--ro ospf:af   iana-rt-types:address-family
  |  |        +--ro ospf:unprotected-routes {fast-reroute}?
  |  |        |  +--ro ospf:af-stats* [af prefix]
  |  |        |     +--ro ospf:af        iana-rt-types:address-family
  |  |        +--ro ospf:local-rib
  |  |        |  +--ro ospf:route* [prefix]
  |  |        |     +--ro ospf:prefix        inet:ip-prefix
  |  |        |     +--ro ospf:next-hops
  |  |        +--ro ospf:statistics
  |  |        +--ro ospf:database
  |  |        |  +--ro ospf:as-scope-lsa-type* [lsa-type]

ISIS model has the following:

module: ietf-routing
  +--rw routing
  |  +--rw control-plane-protocols
  |  |  +--rw control-plane-protocol* [type name]
  |  |     +--rw isis:isis
  |  |        +--rw isis:interfaces
  |  |        |  +--rw isis:interface* [name]
  |  |        |     +--rw isis:name   if:interface-ref
  |  |        +--ro isis:database
  |  |        |  +--ro isis:level-db* [level]
  |  |        |     +--ro isis:level    level-number
  |  |        |     +--ro isis:lsp* [lsp-id]
  |  |        |        +--ro isis:decoded-completed?   boolean
  |  |        |        +--ro isis:raw-data?     yang:hex-string
  |  |        |        +--ro isis:lsp-id        lsp-id
  |  |        +--ro isis:local-rib
  |  |        |  +--ro isis:route* [prefix]
  |  |        |     +--ro isis:prefix       inet:ip-prefix
  |  |        |     +--ro isis:next-hops


This BGP model uses operational state sub-tree mostly from the OpenConfig model, but OpenConfig does not augment ietf-routing and uses separate global tree. If we keep the OpenConfig sub-tree, it would be better to structure the BGP rip as following:

module: ietf-routing
  +--rw routing
  |  +--rw control-plane-protocols
  |  |  +--rw control-plane-protocol* [type name]
           +--rw bgp:bgp
              +--rw global!
              +--rw neighbors
              |  +--rw neighbor* [neighbor-address]
              +--rw peer-groups
                 +--rw peer-group* [peer-group-name]
              +--ro bgp-rib
                 +--ro attr-sets
                 |  +--ro attr-set* [index]
                 |     +--ro index                   uint64
                 +--ro afi-safis
                    +--ro afi-safi* [afi-safi-name]
                       +--ro afi-safi-name       identityref
                       +--ro ipv4-unicast
                       |  +--ro loc-rib
                       |  |  +--ro routes
                       |  |     +--ro route* [prefix origin path-id]
                       +--ro ipv6-unicast
                       |  +--ro loc-rib
                       |  |  +--ro routes
                       |  |     +--ro route* [prefix origin path-id]
                       +--ro ipv4-srte-policy
                       |  +--ro loc-rib
                       |  |  +--ro routes
                       |  +--ro neighbors
                       |     +--ro neighbor* [neighbor-address]
                       +--ro ipv6-srte-policy
                          +--ro loc-rib
                          |  +--ro routes

Thanks,
- Xufeng


2) module ietf-bgp is missing

Is it intentional to remove the main module ietf-bgp? The description says that bgp model augments the ietf-routing, but there is no such an augment statement in the draft. I assume that the augment statement is in the main module ietf-bgp.


That was indeed a cut-and-paste error. The next version of the draft will have the ietf-bgp module.

Thanks.




Thanks,
- Xufeng

On Tue, Feb 26, 2019 at 1:20 PM Mahesh Jethanandani <[e-mail]<mailto:[e-mail]>> wrote:
This update of the draft adds support for:

- augmentation of the Routing Management Model.
- augmentation of the routing policy model
- support for RIB

Comments welcome.

> On Feb 26, 2019, at 10:16 AM, [e-mail]<mailto:[e-mail]> wrote:
>
>
> A New Internet-Draft is available from the on-line Internet-Drafts directories.
> This draft is a work item of the Inter-Domain Routing WG of the IETF.
>
>        Title           : BGP YANG Model for Service Provider Networks
>        Authors         : Keyur Patel
>                          Mahesh Jethanandani
>                          Susan Hares
>       Filename        : draft-ietf-idr-bgp-model-04.txt
>       Pages           : 138
>       Date            : 2019-02-26
>
> Abstract:
>   This document defines a YANG data model for configuring and managing
>   BGP, including protocol, policy, and operational aspects based on
>   data center, carrier and content provider operational requirements.
>
>
> The IETF datatracker status page for this draft is:
> https://datatracker.ietf.org/doc/draft-ietf-idr-bgp-model/
>
> There are also htmlized versions available at:
> https://tools.ietf.org/html/draft-ietf-idr-bgp-model-04
> https://datatracker.ietf.org/doc/html/draft-ietf-idr-bgp-model-04
>
> A diff from the previous version is available at:
> https://www.ietf.org/rfcdiff?url2=draft-ietf-idr-bgp-model-04
>
>
> Please note that it may take a couple of minutes from the time of submission
> until the htmlized version and diff are available at tools.ietf.org<http://tools.ietf.org/>.
>
> Internet-Drafts are also available by anonymous FTP at:
> ftp://ftp.ietf.org/internet-drafts/
>
> _______________________________________________
> Idr mailing list
> [e-mail]<mailto:[e-mail]>
> https://www.ietf.org/mailman/listinfo/idr

Mahesh Jethanandani
[e-mail]<mailto:[e-mail]>



_______________________________________________
Idr mailing list
[e-mail]<mailto:[e-mail]>
https://www.ietf.org/mailman/listinfo/idr

Mahesh Jethanandani
[e-mail]<mailto:[e-mail]>