Re: [Detnet] AD Review of draft-ietf-detnet-ip-oam-09

[Greg Mirsky <[e-mail]>]

Hi Roman,
thank you for your review and comments. Please find my answers and proposed
updates below under the GIM>> tag. Attached, please find the new working
version (Gmail believes that the diff carries a virus ;( ).

Regards,
Greg

On Tue, Oct 31, 2023 at 7:47 PM Roman Danyliw <[e-mail]> wrote:

> Hi!
>
> I performed an AD Review of draft-ietf-detnet-ip-oam-09 and will be taking
> over as the responsible AD on this document to help load-balance work in
> RTG.  Thanks for this document.  My feedback is as follows.
>
> ** Section 3.1
> When the UDP destination port number used by the OAM
>    protocol is one of the assigned by IANA, then the UDP source port can
>    be used to achieve co-routedness of OAM,
>
> Typo? What is “co-routedness”?

GIM>> It is essential for an active OAM method to ensure that the specially
generated test packets traverse the same node and links as the data flow
that is monitored. That is what is referred to as co-routedness between OAM
and the monitored data flow. Also, in MPLS, co-routedness is used to
describe a special case of a bidirectional LSP:

A co-routed bidirectional TE LSP denotes a bidirectional tunnel where the
forward direction LSP and reverse direction LSP must follow the same path,
for example, the same nodes and paths.



>
> ** Section 3.1
>    That correlation between the particular IP OAM
>    protocol session and the monitored IP DetNet flow can be achieved
>    using the DetNet YANG model [I-D.ietf-detnet-yang].
>
> Is this correlation possible by processing the provisioning information
> codified in the YANG?  Is not such a correlation possible by processing
> _any_ provisioning information regardless of the data representation?  I’m
> wondering if this isn’t as simple as “That correlation between the
> particular IP OAM session and the monitored IP DetNet flow can be achieved
> by using DetNet provisioning information (e.g., [I-D.ietf-detnet-yang]).
>
GIM>> Thank you for the proposed text. Accepted.

>
> ** Section 3.2.  Typo.  s/IP active OAM/active IP OAM/?
>
GIM>> Thank you for catching this. Fixed.

>
> ** Section 3.2.
> The amount of operational work mapping IP
>    OAM protocols to the monitored DetNet flow can be reduced by using an
>    IP/UDP tunnel to carry IP test packets.  Then, to ensure that OAM
>    packets traverse the same set of nodes and links, the IP/UDP tunnel
>    must be mapped to the monitored DetNet flow.
>
> -- Using “IP/UDP” tunnel is referenced without citation.
>
GIM>> Thank you. Added the reference to RFC 2003 as follows:
   The amount of operational work mapping IP
   OAM protocols to the monitored DetNet flow can be reduced by using an
   IP/UDP tunnel to carry IP test packets ([RFC2003]).

>
> -- Isn’t this suggesting the opposite of what is said in other places –
> that the measurement and the DetNet packet need to get the same
> treatment/path?
>
GIM>> Because active IP OAM protocols use different UDP port numbers and
even different protocols. Thus, instead of mapping each protocol to the
monitored DetNet flow, all active OAM protocols for the given DetNet flow
are put in a tunnel and that, in turn, is mapped to the DetNet flow. Does
it make sense?

>
> ** Section 3.2
>    [I-D.ietf-detnet-mpls-over-ip-preof] describes how DetNet with MPLS
>    over UDP/IP data plane [RFC9025] can be used to support Packet
>    Replication, Elimination, and Ordering Functions to potentially lower
>    packet loss, improve the probability of on-time packet delivery and
>    ensure in-order packet delivery in IP DetNet's service sub-layer.
>
> -- I don’t understand the link to this PREOF draft and OAM.  What am I
> missing?
>
GIM>> DetNet OAM is required to support the DetNet Service sub-layer
function that is composed of Packet Replication, Elimination, and Order
Preservation functions. The reference to I-D.ietf-detnet-mpls-over-ip-preof
is for future work on the OAM for the DetNet sub-layer. WDYT?

>
> -- I didn’t see a formal treatment of these topics in
> [I-D.ietf-detnet-mpls-over-ip-preof].
>
GIM>> I find that being cursory mentioned in:
   Note: this document focuses on the use of MPLS over UDP/IP
   encapsulation throughout an entire DetNet IP network, making MPLS-
   based DetNet OAM techniques applicable.  Using the described
   encapsulation only for a portion of a DetNet IP network that handles
   the PREOF functionality would complicate OAM.
Would you suggest adding an analogous note in draft-ietf-detnet-ip-oam?

>
> ** Section 6.
>    This document describes the applicability of the existing Fault
>    Management and Performance Monitoring IP OAM protocols, and does not
>    raise any security concerns or issues in addition to ones common to
>    networking or already documented for the referenced DetNet and OAM
>    protocols.
>
> Please provide references to these key DetNet and OAM-related document
> covering the relevant security considerations.
>
GIM>> Thank you for pointing this out to me. Below is the updated text:
NEW TEXT:
   This document describes the applicability of the existing Fault
   Management and Performance Monitoring IP OAM protocols.  It does not
   raise any security concerns or issues in addition to ones common to
   networking or already documented in [RFC0792], [RFC4443], [RFC5880],
   and [RFC8762] for the referenced DetNet and OAM protocols.
I hope that is acceptable.

>
> Thanks,
> Roman
> _______________________________________________
> detnet mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/detnet
>