Hi Les,
Thanks a lot for your comments and suggestions.
Generally I think it's reasonable to discuss requirement and the terms first in
the MPLS/SPRING, as well as in the 6MAN WG I suppose, since the requirement is
not SRv6 specific.
Please see my detailed reply in line with [Yao].
Regards,
Yao
Original
From: LesGinsberg(ginsberg) <[email protected]>
To: 刘尧00165286;[email protected] <[email protected]>;
Date: 2023年10月14日 23:40
Subject: RE: [Lsr] Fw: New Version Notification for draft-liu-lsr-igp-mpd-00.txt
Yao –
A few comments…
While I appreciate that the limits you are defining are different than any of
the limits currently supported by the IGP MSD-Types registry, I think they
still fall conceptually into the same generic bucket i.e., you want to
advertise a limit related to a type of packet encapsulation used for forwarding.
I see no need to create a new registry nor consume a new sub-TLV code point.
The existing IGP MSD registry is very appropriate.
[Yao] Using the existing IGP-MSD sub-TLV is OK from the aspect of
functionality, but it may cause some confusion since MSD is mainly related with
the number of SIDs or labels. This can be discussed further after the
requirement being accepted.
It is not within the purview of the LSR WG to define these new limits. This
needs to be discussed in the MPLS and SPRING WGs, respectively, and preferably
the definitions of the new parameters and discussion of how they can be used is
part of a document in each WG.
The IGP draft then simply becomes a definition of how to advertise values
defined in referenced documents from MPLS/SPRING. The IGP draft is still needed
but must be preceded by the MPLS/SPRING work.
I suggest you put this draft on hold for now.
[Yao] I'll follow your suggestions to discuss these topics in the related WGs
first before the work in IGP. Since the requirement in IPv6/SRv6 is stronger, a
document in this area would be considered first.
The Introduction discusses the parameters as limits to the processing
capability of a given node (or link). If a packet were to arrive with an encap
exceeding this limit it can be assumed that the packet would not be forwarded
correctly.
This makes sense. However, in Section 5 you introduce the phrase “node can read
from an incoming [MPLS/IPv6] packet without performance impact” – which
suggests that a node could indeed process the packet correctly – but it might
be slower than packets received below the limit.
I have difficulty understanding the operational definition of “performance
impact” (Is 10% degradation OK but 50% not OK??) and its use is inconsistent
with the description in the Introduction.
I would expect that when these terms are discussed/defined by the MPLS/SPRING
WGs this confusion will be eliminated – and at that point the IGP draft will
simply reference definitions in other drafts – but I did want to highlight this
inconsistency.
[Yao] The term "without performance impact" comes from the definition of
RLD-MSD in draft-ietf-mpls-mna-fwk. By using this term, part of the meaning is
to process the packet at the fast path /at line rate. In 6MAN, I suppose the
term "at the full forwarding rate" is preferred to express the similar meaning
in draft-ietf-6man-hbh-processing. It would be discussed in the related WGs.
Thanx for listening.
Les
From: Lsr <[email protected]> On Behalf Of [email protected]
Sent: Wednesday, October 11, 2023 2:20 AM
To: [email protected]
Subject: [Lsr] Fw: New Version Notification for draft-liu-lsr-igp-mpd-00.txt
Dear All，
We've submitted a new draft
https://www.ietf.org/archive/id/draft-liu-lsr-igp-mpd-00.html.
This document proposes the concept of Maximum Packet Depth(MPD) to represent
the packet size that a node is able process from an incoming packet, and the
signaling mechanism for MPD in IGP is defined as well.
Welcome your review and comments！
Regards,
Yao
Original
From: [email protected] <[email protected]>
To: 刘尧00165286;沈益明10054407;
Date: 2023年10月10日 15:21
Subject: New Version Notification for draft-liu-lsr-igp-mpd-00.txt
A new version of Internet-Draft draft-liu-lsr-igp-mpd-00.txt has been
successfully submitted by Yao Liu and posted to the
IETF repository.
Name: draft-liu-lsr-igp-mpd
Revision: 00
Title: Signaling Maximum Packet Depth(MPD) using IGP
Date: 2023-10-10
Group: Individual Submission
Pages: 9
URL: https://www.ietf.org/archive/id/draft-liu-lsr-igp-mpd-00.txt
Status: https://datatracker.ietf.org/doc/draft-liu-lsr-igp-mpd/
HTML: https://www.ietf.org/archive/id/draft-liu-lsr-igp-mpd-00.html
HTMLized: https://datatracker.ietf.org/doc/html/draft-liu-lsr-igp-mpd
Abstract:
This document proposes the concept of Maximum Packet Depth(MPD) to
represent the packet size that a node is able process from an
incoming packet, the signaling mechanism for MPD in IGP is also
defined.
The IETF Secretariat_______________________________________________
Lsr mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/lsr
