Re: [Idr] Adoption of ft-zeng-idr-bgp-mtu-extension-01 as IDR WG document?

agreed, I must have misread your draft.

--
Jakob Heitz.

On Nov 17, 2011, at 1:23 AM, "Jie Dong" <jie.dong@huawei.com> wrote:

> Hi Jakob, 
> 
> Thanks for your comments, and please see replies inline with #Jie.
> 
> ________________________________________
> From: idr-bounces@ietf.org [idr-bounces@ietf.org] on behalf of Jakob Heitz [jakob.heitz@ericsson.com]
> Sent: Thursday, November 17, 2011 16:50
> To: John Scudder
> Cc: idr@ietf.org List
> Subject: Re: [Idr] Adoption of ft-zeng-idr-bgp-mtu-extension-01 as IDR  WG      document?
> 
> support.
> 
> For L3VPN option c, BGP contributes one link of an end-to-end LSP. This draft proposes a good way to propagate the PMTU across that link.
> 
> One part I don't understand is in 3.3.c, why does the speaker not pass along the MTU from the previous link? In the L3VPN option c, 3 label case, it sets itself as nexthop. It should preserve the MTU, I think.
> 
> #Jie: When the speaker sets itself as nexthop, this means it would be on the forwarding path of the LSP, and in this case it should compare the received LSP path MTU and "its BGP LSR Link MTU to its nexthop minus 4", and choose the smaller one as its LSP Path MTU, and then propagate this MTU to its upstream BGP peers.
> 
> In some cases, such as a multihop BGP session, or a BGP session between loopback interfaces, it can not determine an MTU. This is of no concern. BGP would simply not generate this MTU community.
> 
> #Jie: In this case BGP could choose the minimum of the link MTU between the two BGP speaker as its BGP LSR link MTU.
> 
> 
> As for data plane/control plane arguments, I think a control protocol that participates in the establishment of a tunnel is the best place to pass any needed properties of that tunnel, including the MTU. Signaling of properties, including MTU belongs in the control plane. That after all is control. The data plane is named "data", because that's all that belongs in it.
> 
> #Jie: Agreed. Path MTU can be considered as a attribute of the LSP path and is reasonable to be advertised using the control plane. And we have done this in other label signaling protocols.
> 
> Regards,
> Jie
> 
> 
> 
> --
> Jakob Heitz.
> 
> 
> On Nov 16, 2011, at 1:04 AM, "John Scudder" <jgs@juniper.net> wrote:
> 
>> Folks,
>> 
>> We have received a request from the authors to adopt ft-zeng-idr-bgp-mtu-extension-01 as an IDR WG document.  Please send your comments to the list.  The deadline for comments is December 5, 2011 at noon EST.
>> 
>> Thanks,
>> 
>> --John
>> _______________________________________________
>> 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