[Bier] Meeting Notes for IETF 117

[Tony Przygienda <[e-mail]>]

I'm posting notes for group review here. will be uploaded in couple days if
no comments

thanks for Hooman and Sandy for taking notes

-- tony


Draft-ietf-bier-mld

  *   Ready for last call
  *   Shepard is assigned waiting for review
  *   Stig
     *   It has not been to the last call yet, it has been around for a
while.
     *   Was waiting for RFC 9279 and some extensions to include sender info
     *   One limitation is if multiple routers are on source LAN we need
some kind of election
        *   One work around is enable this on one of the routers.
        *   Another possibility is run PIM and run PIM DR
        *   Maybe publish this draft with the limitation of single router
or PIM DR on LAN
     *   Poll to wait for a general election draft
        *   8 lets wait and 1 lets push it out now.
        *   Stig seems to be ok with this and for the wait, will talk to
Ice about the general election draft

BIER OAM

  *   Update on progress Greg Mirsky
  *   BIER OAM requirement, ping ,path mtu discovery, bfd performance
measurement
  *   Early reviews from security and Routing directors are in
     *   Good discussions and updates base on these reviews
  *   Some TLV update and IANA section update
  *   Nokia has implementation for OAM PING and notes will be provided
* Greg asked for WG LC for the req draft
  *   On BIER BFD
     *   Two use cases RFC 8562 BFERs can Monitor the BFIR and RFC 8563
Active Tails
     *   Routing Dir review seems to be positive
  *   Hybrid performance measurement in BIER
     *   Active measurements in Multicast environment, provide comments
     *   Tony
        *   Call for adoption?
        *   Greg, not yet as there are some scenarios that have technical
issues with Hybrid Performance Measurement draft in general
        *   Tony, is BIER going to just piggy back on this?
           *   Think applicability, deployment scenarios for BIER
        *   Greg do we need specific BIER BIT MASK and TLV for this
Performance Measurement or L2/L3 TLVs will do?
        *   Ready for review and comments

BIER in 6

  *   Sandy Zhang
     *   BIER in (none MPLS) IPv6 networks
        *   The none mpls encap
        *   Can be used for IPv6 tunneling or even hop by hop
        *   For hop by hop the ipv6 header is optional
     *   Brief introduction
        *   Existing BIER procedures are valid for BIER in 6 as well
     *   All underlay drafts are passed last call
        *   Requesting for last call and more comments
        *   9 (or 10) people are thinking it is ready for LC no opposition

ISIS extensions for BIER-TE

  *   Presenter Huaimo
  *   Removed padding and extra text
  *   Added sub-TLV and sub-sub-TLVs
     *   New sub-TLVs are in the draft read it! (in total 3)
  *   Comments
     *   Jeffrey, 3 drafts for ISIS and OSPF and SPPF3 for bit positions,
and another draft for BIFT ID signaling, we should have only one draft for
bit position and another draft for bit signaling for BGP. BFID should be
added part of the same draft
     *   Huaimo, The reason we have 3 drafts because we just follow the
IPv4 example and extensions, ISIS, OSPF and OSPF3
     *   Jeffrey, I think I missed that comment in the WG, in early days it
made sense but now as BIER is mature it is just incremental, so at this
point it makes sense to have a single draft for ISIS, OSPF, and OSPFv3 and
a separate for BGP
     *   Huaimo, open but I think ISIS one draft and OSPF another draft as
the coding is different.
     *   Jeffrey, just pointing my preference but up to WG
     *   Greg, original ISIS, OSPF were the foundation but now it is
similar extensions. So Greg is leaning toward Jeffrey single doc. But lets
go to WG.

OSPF, OSPFv3 extensions for BIER-TE

  *   Presenter Huaimo

  *   Changes to TLV and SUB-TLVs and SUB-SUB-TLVs
  *   Read the draft to get the update on the changes.
  *   Comments?

BIER-TE for Broadcast Link

  *   Presenter Huaimo

  *   Sandy
     *   I don't think this is a true problem, this can be avoid via a
controller. The controller won't program the wrong bit map
     *   Huaimo, yes you can use a controller to avoid the problem but
seems to be a problem, there is not information to indicate the remote
node. Each node has capability to process the local bits positions and not
beyond
     *   Sandy but if you read the RFC9262 to create a multicast tree there
might be a corner case that there is duplicate. Consider carefully.
     *   Lets discuss farther in the list

BIER Extension Header

  *   Jeffrey presenter
  *   Cases to extend BIER functions and need the BIER Extension header as
an example Fragmentation and Reassembly (might not have same encoding but
are good examples)
  *   There might be multiple other drafts that are trying to solve this as
well as an example in MPLS WG
     *   Andrew, debate on MPLS side ISD vs PSD if we need for alignment we
need to follow these WG closely (interim was held in the meantime)
     *   Tony, we need to make an effort to align with MPLS
     *   Greg Mirsky, the status in MPLS WG is still open, the discussion
is around what use cases do we that can't be solved using ISD, until now
all use cases can be solved using ISD. Is BIER one that can't be solved via
ISD?
     *   Jeffrey prefers to go with PSD solution and not ISD so it is
aligned with IPv6 encoding
     *   Greg, valid concerns the decision that BIER group takes is for
BIER consideration and BIER is not forced to follow
     *   Sandy, BIER should consider the extension header as the PSD in
MPLS has no relation with BIER
     *   Tony, read the MPLS requirements and see what is related to BIER
and make a decision
  *   Seeking comments and suggestions and BIER is looking closely as MPLS
PSD, if they don't conclude we would follow IPv6 and conclude in BIER side.
     *   Jeffrey, the only concrete extension is BIER OAM, security,
fragmentation  etc.. to follow
     *

Discussion of BIER interop events

Jeffrey Zhang: there are some off-line discussion about BIER interop events
as part of hackathon in Prague. I will send the email to public BIER
mailing list.
Greg Shepherd: who do we have confirmed so far?
Jeffrey Zhang: Nokia, Juniper, Huawei and ZTE have the indication but
that’s just unofficial. I forgot to send email to Cisco.
Greg Shepherd: don’t forget P4 implementation.
Hooman Bidgoli: we may need to make sure how do we make the devices
connection., because some implementations using simulator and some
implementations like Nokia using the actual hardware.
Jeffrey Zhang: even in simulator there may need hardware though the
connection by simulator may be easier. They all are the details we need to
figure out. we should have the discussion ahead.
Greg Shepherd (without chair’s hat): let me know if you need an independent
third party to sit in the middle.