Hi Paolo,
Sorry for delayed reply too - was on vacation. :(Answers are inline, BKH>
Thanks.
SY,Boris
On Tuesday, April 29, 2025 at 12:19:58 PM GMT+3, Paolo Volpato
<[email protected]> wrote:
Hi Boris,
Thank you for your comments and apologies for the belated answer.
Please see our answer inline, tagged as [PV].
Best regards
Paolo (on behalf of the co-authors)
From: Boris Hassanov <[email protected]>
Sent: Friday, April 11, 2025 8:25 PM
To: [email protected]
Cc: SPRING WG List <[email protected]>; SRv6 Operations Discussion List
<[email protected]>
Subject: [bmwg] Re: I-D Action: draft-ietf-bmwg-sr-bench-meth-03.txt
Hi authors and all,
As promised during the IETF, I would like to provide additional comments about
the draft.
Items 2 and 3 both say about a SR Policy, similarly 4.2 that also pays
attention for the long SID list. However forwarding performance tests (6) do
not explicitly mention: which method DUT should use for 'SID stack
construction' (as per 4.2).
[PV] We can mention in section 6 the possibility to test more construction
methods and report the results as specified in section 5 on the reporting
format (where a parameter is already defined).
BKH> sounds good for me.
Thus I would suggest to separate a SR Policy on a source edge node in a
specific subcase (abstracting from the way to provision it - could be any) and
define more detailed testing methodology, i.e. defining the way how to steer
traffic into SR Policy, measuring the performance and LB ratio starting from 1
CP, 1 SL with 2 SIDs, then up to MSD, testing ECMP/wECMP amongst NxSLs (i.e.
16, 32, 64,128) with different number of SIDs and frame sizes (4.3). The goal
is to compare SR Policy performance with other methods, check ECMP/wECMP limits
and correct SR Policy work overall.
[PV] Do you mean to test a single SR policy on a source edge node with more
construction methods (to check, for example, what is the more performant) or to
test more SR policies at the same time (to check the overall ECMP performance)?
BKH> I meant testing one SR Policy with 1 active CP but with variable SLs
number, SIDs, weights. I can unicast more detailed test description to you.
SY,
Boris
On Monday, March 3, 2025 at 11:53:42 AM GMT+3, <[email protected]> wrote:
Internet-Draft draft-ietf-bmwg-sr-bench-meth-03.txt is now available. It is a
work item of the Benchmarking Methodology (BMWG) WG of the IETF.
Title: Benchmarking Methodology for Segment Routing
Authors: Giuseppe Fioccola
Eduard Vasilenko
Paolo Volpato
Luis Miguel Contreras
Bruno Decraene
Name: draft-ietf-bmwg-sr-bench-meth-03.txt
Pages: 28
Dates: 2025-03-03
Abstract:
This document defines a methodology for benchmarking Segment Routing
(SR) performance for Segment Routing over IPv6 (SRv6) and MPLS (SR-
MPLS). It builds upon RFC 2544, RFC 5180, RFC 5695 and RFC 8402.
The IETF datatracker status page for this Internet-Draft is:
https://datatracker.ietf.org/doc/draft-ietf-bmwg-sr-bench-meth/
There is also an HTMLized version available at:
https://datatracker.ietf.org/doc/html/draft-ietf-bmwg-sr-bench-meth-03
A diff from the previous version is available at:
https://author-tools.ietf.org/iddiff?url2=draft-ietf-bmwg-sr-bench-meth-03
Internet-Drafts are also available by rsync at:
rsync.ietf.org::internet-drafts
_______________________________________________
bmwg mailing list [email protected]
To unsubscribe send an email [email protected]
_______________________________________________
spring mailing list -- [email protected]
To unsubscribe send an email to [email protected]
_______________________________________________
spring mailing list -- [email protected]
To unsubscribe send an email to [email protected]
