Skip to main content

Minutes for GROW at IETF-94
minutes-94-grow-1

Meeting Minutes Global Routing Operations (grow) WG
Date and time 2015-11-02 06:20
Title Minutes for GROW at IETF-94
State Active
Other versions plain text
Last updated 2015-11-01

minutes-94-grow-1
Notes from the IETF94 GROW meeting in Yokohama, Japan.

Chairs: Christopher Morrow & Peter Schoenmaker
Jabber: Colin Petrie, Notes: Job Snijders

#0 Thomas King - draft-ymbk-grow-blackholing

    King offers background information on why a standardized blackhole
    BGP community is useful. Presents on different methods current
    implemented by various organisations: each one uses different
    communities for the same function.

    65535:666 is (still) the desired identifier for the well-known community.

    As far as King knows there are no outstanding comments/remarks.

    Morrow indicates that the draft will probably soon proceed to next-call.

#1 Colin Petrie - MRT/BGP Add-Paths draft-petrie-grow-mrt-add-path

    In IDR people work on draft-ietf-idr-add-paths. This brings new NLRI
    encoding for BGP messages, negotiated via capabilities.

    Problem statement: Current MRT parsers cannot know how many bttes
    should be in the NLRI. You can reverse engineer how many bytes it
    should read for that field, if you scan back (possibly several
    weeks) to the negotiated capabilities in the OPEN messages.

    Proposal: add new SubType codes to signal to the parser how to
    interpret the NLRI and MP_NLRI fields.

    Peter asks the room whether there is support to adopt the document
    as GROW working group doc. The room is positive about adopting, no
    objections.

#2 Jared Mauch - draft-mauch-bgp-reject

    "An EBGP speaking device MUST NOT advertise routes without
    configured policy". Some router vendors already conform to the
    draft: IOS XR & JunOS.

    Mauch is looking for home/working group for the document.

    Manning suggests that the document should not limit itself to eBGP
    but to any and all BGP. The room near unanimously reacts with "no".

    Volk (DT) appreciates the draft, and wonders what other advice
    implementors need to create proper: e.g. sementics of well-known
    communities is often not honored by default.

    Schoenmaker asks the room whether there is support to make the
    document a working group document. There is strong support in the
    room to take this on.

End of meeting: 15:50 local time.