Re: [Lsr] Flooding Topology Computation Algorithm - draft-cc-lsr-flooding-reduction-08 Working Group Adoption Call

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

Hi Huaimo,

From: Huaimo Chen <[e-mail]>
Date: Friday, June 5, 2020 at 9:03 PM
To: Acee Lindem <[e-mail]>, "[e-mail]" <[e-mail]>
Subject: Re: [Lsr] Flooding Topology Computation Algorithm - draft-cc-lsr-flooding-reduction-08 Working Group Adoption Call

Hi Acee,

    Thank you very much for your valuable comments.
    My answers/explanations are inline below.

Best Regards,
Huaimo
________________________________
From: Lsr <[e-mail]> on behalf of Acee Lindem (acee) <acee=[e-mail]>
Sent: Friday, June 5, 2020 12:52 PM
To: [e-mail] <[e-mail]>
Subject: Re: [Lsr] Flooding Topology Computation Algorithm - draft-cc-lsr-flooding-reduction-08 Working Group Adoption Call


Speaking as WG Member:



Hi Authors,



I have a couple technical comments on the draft.



1.      The algorithm only computes a single graph with single path to each node. While I-D.ietf-lsr-dynamic-flooding failure and computation of an updated flooding topology, many of the other algorithms (both draft and proprietary) attempt to compute two or more paths to each node.
[HC]: The flooding topology (or say a graph) computed by the algorithm has at least two paths to each node. At step 4 of the algorithm on page 5 of the draft, for each node B that has only one path to it on the flooding topology (i.e., the Degree of node B is one) , another path is added to it. Thus the final flooding topology (or say graph) computed by the algorithm has  two or more paths to every node.

Ok – I see this now.





2.      Since this algorithm is slated for use as a distributed algorithm, all routers in the area must use the same value for MaxD or they may compute different flooding topologies. This is cannot be guaranteed by a single algorithm IANA value unless we allow some number of algorithm specific parameters to be advertised by the area leader (which might not be a bad idea).
[HC]: You are right. All the routers must use the same value for MaxD. We will update the draft for this accordingly. For now, MaxD has an initial value of 3 (i.e., every router uses this initial value of 3 for MaxD).  The algorithm increases MaxD by one and restarts to compute the flooding topology with this new MaxD if it can not compute a flooding topology with the old MaxD. After some iterations in some cases, the flooding topology is computed. It is a good idea to pass some parameters from the area leader to every router when the leader advertises the algorithm number to every router.  MaxD can be one of the parameters.

3.       Similarly, the constraints in section 4.2 must be applied uniformly and different constraints would require new IANA algorithm allocation.
[HC]: You are right. All the routers must use the same constraints.  A new (or different) set of constraints need a new algorithm number. We will update the draft for this accordingly.



I see you’ve updated this in the latest.



Thanks,

Acee



Thanks,
Acee



From: Lsr <[e-mail]> on behalf of "Acee Lindem (acee)" <acee=[e-mail]>
Date: Friday, May 15, 2020 at 3:40 PM
To: "[e-mail]" <[e-mail]>
Subject: [Lsr] Flooding Topology Computation Algorithm - draft-cc-lsr-flooding-reduction-08 Working Group Adoption Call



This begins a 3 week (due to holidays) WG adoption call for the “Flooding Topology Computation Algorithm” draft. Please issue your support or objection to this list by 11:59 PM, UTC on June 5th, 2020. Here is a URL for your convenience.



https://datatracker.ietf.org/doc/draft-cc-lsr-flooding-reduction/<https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdatatracker.ietf.org%2Fdoc%2Fdraft-cc-lsr-flooding-reduction%2F&data=02%7C01%7Chuaimo.chen%40futurewei.com%7C1a7e82ec42fa40a9ad8e08d80970ea4c%7C0fee8ff2a3b240189c753a1d5591fedc%7C1%7C0%7C637269727959858751&sdata=eG3iR27JyxGL6qCsqlMow47twBhMpf0%2BsgJo6Q9hBJg%3D&reserved=0>



Thanks,
Acee