In EVPN, as Patrice described, the structure is:

 

      |     +--rw bgp-parameters

      |     |  +--rw common

      |     |     +--rw rd-rt* [route-distinguisher]

      |     |        +--rw route-distinguisher    string

      |     |        +--rw vpn-target* [rt-value]

      |     |           +--rw rt-value    string

      |     |           +--rw rt-type     bgp-rt-type

 

In L2VPN, the structure is:

             +--ro bgp-auto-discovery

             |  +--ro route-distinguisher?   string

             |  +--ro vpn-target* [rt-value]

             |  |  +--ro rt-value    string

             |  |  +--ro rt-type     bgp-rt-type

 

In L3VPN, the current structure is:

      +--rw route-distinguisher

      |  +--rw config

      |  |  +--rw rd?   string

      +--rw ipv4

      |  +--rw unicast

      |     +--rw route-targets

      |     |  +--rw config

      |     |  |  +--rw rts* [rt]

      |     |  |  |  +--rw rt         string

      |     |  |  |  +--rw rt-type?   Enumeration

      +--rw ipv6

         +--rw unicast

            +--rw route-targets

            |  +--rw config

            |  |  +--rw rts* [rt]

            |  |  |  +--rw rt         string

            |  |  |  +--rw rt-type?   enumeration

 

Hi Dhanendra and All, 

Are we ok to move the route targets section out of the AF specific location
to where RD is specified? If so, we can define the following common
grouping:

 

             |  +--ro route-distinguisher?   string

             |  +--ro vpn-target* [rt-value]

             |  |  +--ro rt-value    string

             |  |  +--ro rt-type     bgp-rt-type

 

Otherwise, we can only define a grouping without the RD:

 

             |  +--ro vpn-target* [rt-value]

             |  |  +--ro rt-value    string

             |  |  +--ro rt-type     bgp-rt-type

Thanks,

- Xufeng

 

 

From: BESS [mailto:[email protected]] On Behalf Of Acee Lindem (acee)
Sent: Saturday, February 11, 2017 12:46 PM
To: Patrice Brissette (pbrisset) <[email protected]>; Jeff Tantsura
<[email protected]>; Giles Heron <[email protected]>
Cc: [email protected]; Dhanendra Jain (dhjain)
<[email protected]>; [email protected]; Shah, Himanshu <[email protected]>
Subject: Re: [bess] BGP common parameter Yang module

 

Given that there is no paucity of authors and contributors on these three
BESS YANG models, I'd hope that one of them could provide a suggested common
grouping. For now, I've added the route-target-type type on which there
seems to be consensus. 

 

Thanks,

Acee  

 

From: "Patrice Brissette (pbrisset)" <[email protected]
<mailto:[email protected]> >
Date: Saturday, February 11, 2017 at 8:23 AM
To: Jeff Tantsura <[email protected] <mailto:[email protected]>
>, Acee Lindem <[email protected] <mailto:[email protected]> >, Giles Heron
<[email protected] <mailto:[email protected]> >
Cc: Himanshu Shah <[email protected] <mailto:[email protected]> >,
"[email protected] <mailto:[email protected]> " <[email protected]
<mailto:[email protected]> >, "Dhanendra Jain (dhjain)" <[email protected]
<mailto:[email protected]> >
Subject: Re: [bess] BGP common parameter Yang module

 

Hi Folks,

 

Same here. Can we do something about it?  And agree, all 3 VPN models should
have the same commonality.

 

Regards,

Patrice Brissette

 

From: Jeff Tantsura <[email protected]
<mailto:[email protected]> >
Date: Friday, February 10, 2017 at 2:43 PM
To: "Acee Lindem (acee)" <[email protected] <mailto:[email protected]> >, Giles
Heron <[email protected] <mailto:[email protected]> >
Cc: Patrice Brissette <[email protected] <mailto:[email protected]> >,
"Shah, Himanshu" <[email protected] <mailto:[email protected]> >, "[email protected]
<mailto:[email protected]> " <[email protected] <mailto:[email protected]> >, "Dhanendra
Jain (dhjain)" <[email protected] <mailto:[email protected]> >
Subject: Re: [bess] BGP common parameter Yang module

 

I'd prefer common grouping indraft-ietf-rtgwg-routing-types and references
from any other model using it 

 

 

Cheers,

Jeff

 

 

From: BESS <[email protected] <mailto:[email protected]> > on behalf
of "Acee Lindem (acee)" <[email protected] <mailto:[email protected]> >
Date: Friday, February 10, 2017 at 08:42
To: Giles Heron <[email protected] <mailto:[email protected]> >
Cc: "Patrice Brissette (pbrisset)" <[email protected]
<mailto:[email protected]> >, "Shah, Himanshu" <[email protected]
<mailto:[email protected]> >, "[email protected] <mailto:[email protected]> "
<[email protected] <mailto:[email protected]> >, "Dhanendra Jain (dhjain)"
<[email protected] <mailto:[email protected]> >
Subject: Re: [bess] BGP common parameter Yang module

 

Hi Giles,

I will add the route-target-type type (enum of import, export, both) but for
a general grouping, it appears there are some discrepancies between the 3
models. Assuming the types: route-discriminator, route-target, and
route-target-type, can you provide a consensus grouping that all the models
would use? 

Thanks,

Acee 

 

From: Giles Heron <[email protected] <mailto:[email protected]> >
Date: Friday, February 10, 2017 at 11:18 AM
To: Acee Lindem <[email protected] <mailto:[email protected]> >
Cc: "Patrice Brissette (pbrisset)" <[email protected]
<mailto:[email protected]> >, "[email protected] <mailto:[email protected]> "
<[email protected] <mailto:[email protected]> >, Himanshu Shah <[email protected]
<mailto:[email protected]> >, "Dhanendra Jain (dhjain)" <[email protected]
<mailto:[email protected]> >
Subject: Re: [bess] BGP common parameter Yang module

 

Hi Acee, 

 

In general seems that for any BGP VPN (L2 or L3) you have an RD plus a list
of RTs (which can be import, export or both) - so I'd prefer that to be
defined in a shared grouping (more or less as per the structure Patrice gave
below) than to force each model to redefine it.

 

Giles

 

On 10 Feb 2017, at 14:51, Acee Lindem (acee) <[email protected]
<mailto:[email protected]> > wrote:

 

Hi Patrice - we are working fervently on a common IETF routing types model.
We have both route-target and router-distinguisher types defined there. The
work is being done in the Routing WG. Our intension is to accelerate
standardization so it doesn't hold up standardization of the importing
modules. Please comment as to whether you think this meets BESS
requirements. 

 

 <https://www.ietf.org/id/draft-ietf-rtgwg-routing-types-00.txt>
https://www.ietf.org/id/draft-ietf-rtgwg-routing-types-00.txt

 

Thanks,

Acee 

P.S. We plan an update next week but the RD and RT definitions have not
changed. 

 

 

 

From: BESS < <mailto:[email protected]> [email protected]> on behalf
of "Patrice Brissette (pbrisset)" < <mailto:[email protected]>
[email protected]>
Date: Friday, February 10, 2017 at 9:26 AM
To: " <mailto:[email protected]> [email protected]" < <mailto:[email protected]>
[email protected]>
Cc: "Dhanendra Jain (dhjain)" < <mailto:[email protected]> [email protected]>,
Himanshu Shah < <mailto:[email protected]> [email protected]>
Subject: [bess] BGP common parameter Yang module

 

Folks,

 

As part of EVPN, L2VPn and L3VPN Yang model, there is a "module" common to
all 3 Yang models.

 

      |     +--rw bgp-parameters

      |     |  +--rw common

      |     |     +--rw rd-rt* [route-distinguisher]

      |     |        +--rw route-distinguisher    string

      |     |        +--rw vpn-target* [rt-value]

      |     |           +--rw rt-value    string

      |     |           +--rw rt-type     bgp-rt-type

 

 

It will be interesting to create a common BGP parameter Yang module as shown
above. I think it just makes sense.

However, there is a minor challenge; that module require a home (a draft).

I'm looking for feedback about the best place/draft for such a module.

 

Thanks for your help.

Regards,

Patrice Brissette

_______________________________________________
BESS mailing list
 <mailto:[email protected]> [email protected]
 <https://www.ietf.org/mailman/listinfo/bess>
https://www.ietf.org/mailman/listinfo/bess

 

_______________________________________________ BESS mailing list
[email protected] <mailto:[email protected]>
https://www.ietf.org/mailman/listinfo/bess 

_______________________________________________
BESS mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/bess

Reply via email to