Re: [ippm] New Draft draft-gandhi-spring-udp-pm

["Rakesh Gandhi (rgandhi)" <[e-mail]>]

Thanks Mach for the comments. We will add following TLV in the next revision of the draft.


3.1.2.1.  Block Number TLV

   The Loss Measurement using Alternate-Marking method defined in
   [RFC8321] requires to identify the Block Number (colour) of the
   traffic counters carried by the probe query and response messages.
   Probe query and response messages specified in [RFC6374] for Loss
   Measurement do not define any means to carry the Block Number.

   [RFC6374] defines probe query and response messages that can include
   one or more optional TLVs.  New TLV Type (value TBA8) is defined in
   this document to carry Block Number (32-bit) for the traffic counters
   in the probe query and response messages for loss measurement.  The
   format of the Block Number TLV is shown in Figure 11:

    0                   1                   2                   3
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    |   Type TBA7   |    Length     |      Reserved                 |
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    |                       Block Number                            |
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

                    Figure 11: Block Number TLV

   The Block Number TLV is optional.  The PM querier node SHOULD only
   insert one Block Number TLV in the probe query message and the
   responder node in the probe response message SHOULD return the first
   Block Number TLV from the probe query messages and ignore other Block
   Number TLVs if present.  In probe query and response messages, the
   counters MUST belong to the same Block Number.


Thanks,
Rakesh



From: Mach Chen <[e-mail]>
Date: Monday, July 23, 2018 at 11:31 PM
To: "=SMTP:rgandhi@cisco. com" <[e-mail]>
Cc: "[e-mail]" <[e-mail]>, spring <[e-mail]>
Subject: RE: New Draft draft-gandhi-spring-udp-pm

Hi Authors,

The draft-gandhi-spring-udp-pm-01 defines a new C flag as following:

3.1.2.1.  Loss Measurement Flags

   An LM message carries Data Format Flags (DFlags) as defined in
   [RFC6374].  New Flag is defined in this document for Color (C) in the
   DFlags field as follows.

                              +-+-+-+-+
                              |X|B|C|0|
                              +-+-+-+-+

                          Data Format Flags

   The Flag C indicates the Color of the counters in the LM probe
   message [RFC6374] when using Alternate-Marking method defined in
   [RFC8321].
-------------

As defined in Section 4.2 of [RFC8321], could you consider to define more than one flag or a TLV to carry Block number instead?

Best regards,
Mach

From: ippm [mailto:[e-mail]] On Behalf Of Rakesh Gandhi (rgandhi)
Sent: Tuesday, July 17, 2018 9:57 PM
To: [e-mail]
Subject: [ippm] New Draft draft-gandhi-spring-udp-pm

Hi WG,

We like to introduce following new draft that was presented to SPRING WG yesterday.

This draft defines IP/UDP path for sending probe query messages for delay and loss measurement that is agnostics to data plane (SR-MPLS/SRv6/IP) and does not require to bootstrap PM session.


https://datatracker.ietf.org/doc/draft-gandhi-spring-udp-pm/

You may find presentation in the following package (it is the second draft).

https://datatracker.ietf.org/meeting/102/materials/slides-102-spring-13-performance-measurement-in-sr-networks-00

We welcome your comments and suggestions.

Thanks,
Rakesh (On behalf of authors and contributors)