Minutes IETF104: mpls
minutes-104-mpls-01

MPLS session - Wednesday, March 27, 2019 - 11:20-13:20 Morning session II
No.        I-D        Version        Start Time        End Time        Duration
(mins)        Presenter         Slides Received 1        Agenda bashing, WG
status reports        -        11:20:00        11:35:00        15        Chairs
2        draft-nslag-mpls-deprecate-md5        4        11:36:00       
11:44:00        8        Andrew Malis        Y 3       
draft-andersson-mpls-spl-terminology        1        11:45:00        11:55:00  
     10        Adrian Farrel        Y 4       
draft-hu-mpls-sr-inter-domain-use-cases        1        11:56:00       
12:06:00        10        Quan        Y 5       
draft-zheng-mpls-lsp-ping-yang-cfg        10        12:07:00        12:22:00   
    15        Greg MIRSKY        Y 6       
draft-nainar-mpls-rfc8287-len-clarification        0        12:23:00       
12:33:00        10        Nagendra Kumar        Y 7       
draft-chen-mpls-cqf-lsp-dp        0        12:34:00        12:42:00        8   
    Zhe Chen        Y 8        draft-cheng-mpls-inband-pm-encapsulation       
0        12:43:00        12:51:00        8        Xiao Min        Y 9       
draft-kumar-mpls-mint        0        12:51:00        13:01:00        10       
Jai Kumar        Y

1) Agenda (Loa):

Misref (draft-ietf-mpls-spring-entropy-label) document:
    Adrian:     Waiting on the base SR draft.
                    That completed LC and waiting for the sheppard write-up to
                    move forward

2)        draft-nslag-mpls-deprecate-md5        4        11:36:00       
11:44:00        8        Andrew Malis        Y

Rajiv Asati:
    aware of 10 (at least) network that use LDP with MD5
    Would they care to replace MD5 with AO - not sure probably no
    Maybe draft would have clear recommendation on what to use

    Adrian :
        use of MD5 as security
        MD5 is not secure to be called security
        Nothing happened so far - so why care, is this enough to call it secure?

Loa:
    MD5 is there in networks
    do they know/care if it is secure enough
    what are the threats and are they addressed by MD5, by AO?

Deborah:
    agree with Adrian.
    It is important to discuss this

3)        draft-andersson-mpls-spl-terminology        1        11:45:00       
11:55:00        10        Adrian Farrel        Y

Julien Muric:
    Does it mean we can have ebSPL?

4)        draft-hu-mpls-sr-inter-domain-use-cases        1        11:56:00     
  12:06:00        10        Quan        Y

Loa:
    SPRING charter says that it works within 1 trust domain (with some
    exception) The work presented is for cross-domains, -> Deborah should
    charter for SPRING change May need some discussion on trust domains
    assumption can be made that SR domain and MPLS-TP domain

Deborah:
    with care of the word "domain" what it means
    Work with the chairs of SR and this is still an early document
    suggest to cc the SR list to engage/feedback

{...}/Juniper:
    what do you mean by domain? IGP area? AS? technoliogy domain?

Charle Halpern
        To interwork between SR-MPLS and MPLS-TP
        Suggest to drop the inter-domain argument to simplify the discussion

Greg Mirstky:
    clarify how the name is used in the context of this proposal (as Debora
    said) Suggest aligning with interpretation with SR-MPLS terminolgy We'll
    need to clarify mapping of SR-domain and this usecase

Adrian:
    SPRING used "SR domain" as a trust domain

5)        draft-zheng-mpls-lsp-ping-yang-cfg        10        12:07:00       
12:22:00        15        Greg MIRSKY        Y Loa:
    Started a WGAP, but found issues.
    Closed WGAP and asked Greg to work with reviewers to address comments and
    will restart the WGAP later

Tarek/Nagendra: the RSVP IPv4 LSP FEC and IPv6 LSP FEC is defined and used in
OAM

6)        draft-nainar-mpls-rfc8287-len-clarification        0        12:23:00 
      12:33:00        10        Nagendra Kumar        Y

Loa: procedural. You need to send a request for WGAP (for book keepoing)

7)        draft-chen-mpls-cqf-lsp-dp        0        12:34:00        12:42:00  
     8        Zhe Chen        Y

Rajiv:
    how upstream router learns ...

Greg:
    For DETNET - even for local behavior -- document is useful to clarify
    xx

Title mention CQF (IEEE) and then large-scale
A: this draft target large scale and LDN .. maybe terms are not correct

Loa:
    there are some questions that we're asking clarification, so lets use the
    list to clarify

8)        draft-cheng-mpls-inband-pm-encapsulation        0        12:43:00    
   12:51:00        8        Xiao Min        Y Qs: none Loa: authors need to
push the discussion on the list

9)        draft-kumar-mpls-mint        0        12:51:00        13:01:00       
10        Jai Kumar        Y Greg:
    We had the discussion when original proposal when iOAM and this may be
    misleading Active OAM can be in-band In MPLS, with EL then they are in-band
    in-band: follow same nodes/links
A: open to changes in the name of the draft

Greg:
    2 or 3 proposals in IPPM - not dataplane specific
    on-path telemetry collection
    instead of using label - why not use GACH type
A:GAch: was considered but there was issues - will add to appendix
    Have you looked at the applicability of this to SR-MPLS
    May have multiple insertions due to

Adrian:
    Please take out the IANA allocated value and add TBD

Nagendra:
    what happen on the egress
    A: there are 2 ways to handle it - may be ok for flexible NPU

Greg:
    I strongly encourge to read the other propsals

Loa: expect authors to pushing discussion on the list