Re: [Idr] draft-head-idr-bgp-ls-isis-fr-01 - WG adoption call (6/6 to 6/20)

["Acee Lindem (acee)" <[e-mail]>]

Hi Robert,

From: Robert Raszuk <[e-mail]>
Date: Tuesday, June 21, 2022 at 7:44 PM
To: Acee Lindem <[e-mail]>
Cc: "Les Ginsberg (ginsberg)" <ginsberg=[e-mail]>, IDR List <[e-mail]>, Susan Hares <[e-mail]>
Subject: Re: [Idr] draft-head-idr-bgp-ls-isis-fr-01 - WG adoption call (6/6 to 6/20)

Hi Acee,

I am afraid the OSPF transport instance draft you quote serves a completely different purpose.

I believe your perspective is limited by the solution you have in mind. This would work perfectly for LSDB dissemination (note that there is a precedent here). It could also be extended to IS-IS LSPs (with a little invention).

IMP (IGP Monitoring Protocol) I suggested would be an analogy to BMP (BGP Monitoring Protocol) unicasting either received flooded LSPs or LSAs or content of LSDB over TCP or QUIC to some central oracle.

I am not aware such a proposal was ever discussed in LSR WG or its predecessors WGs.

I feel this protocol would be better discussed in the RTG WG.

Thanks,
Acee

Many thx,
Robert

PS. And there were and are many real world use cases for such. Till now folks who needed to listen to link state flooding were always asked to set up a passive session. But we do see this is not flexible enough. A p2p protocol is needed. BGP-LS is a living proof of it. In fact IMP could leverage the exact same encoding as currently sent over BGP-LS SAFI.






On Wed, Jun 22, 2022 at 1:35 AM Acee Lindem (acee) <[e-mail]<mailto:[e-mail]>> wrote:


From: Idr <[e-mail]<mailto:[e-mail]>> on behalf of Robert Raszuk <[e-mail]<mailto:[e-mail]>>
Date: Tuesday, June 21, 2022 at 7:23 PM
To: "Les Ginsberg (ginsberg)" <ginsberg=[e-mail]<mailto:[e-mail]>>
Cc: IDR List <[e-mail]<mailto:[e-mail]>>, Susan Hares <[e-mail]<mailto:[e-mail]>>
Subject: Re: [Idr] draft-head-idr-bgp-ls-isis-fr-01 - WG adoption call (6/6 to 6/20)

Dear Les,

I have two final questions on this:

on such a mundane topic.

Would IDR keeping their WG policy in place and mandating two interoperable implementations and decent industry support make this topic a little bit less mundane ?

If someone actually has an alternative proposal with some detail – that would be interesting.

This proposal to send LSDB across N nodes should come from and within LSR WG as we are talking about LSR protocols. Why IDR is even cc-ed and LSR WG is not beats me.

Why can't LSR WG not come with IMP - IGP Monitoring Protocol ?

We already have the capability in OSPF – See section 3.3 of https://www.ietf.org/archive/id/draft-ietf-lsr-ospf-transport-instance-02.txt  We just need more support…

Thanks,
Acee



Thx,
R.

PS. Actually the way I read Jeff's note was very positive and extended help to LSR WG to make the process seamless. But it seems some folks apparently think this is all done - do not waste our time - You IDR-ers take what we produce in ISIS and OSPF to BGP forever. That's pretty cool indeed :)  And another LSR WG member and ex-chair says if you do not take it we will squat codepoints and do it anyway ... so you have no choice. Bravo !