Acee -
The section in the draft is only some suggestions - it isn’t normative and
clearly states that.
I appreciate you promoting LSVR work, but the draft you reference is
experimental and not yet achieved WG last call.
Not clear to me that a reference is needed at this time.
Consider this a "gentle" pushback. 😊
Les
> -----Original Message-----
> From: Acee Lindem <[email protected]>
> Sent: Friday, March 7, 2025 3:55 AM
> To: Les Ginsberg (ginsberg) <[email protected]>
> Cc: Paul Congdon <[email protected]>;
> [email protected]; lsr <[email protected]>; draft-glctgp-lsr-l2-bundle-
> [email protected]
> Subject: Re: WG Adoption Poll for "Advertisement of Remote Interface
> Identifiers for Layer 2 Bundle Members" -draft-glctgp-lsr-l2-bundle-member-
> remote-id-02
>
> Hi Les,
>
> > On Mar 6, 2025, at 11:48 PM, Les Ginsberg (ginsberg)
> <[email protected]<mailto:[email protected]>> wrote:
> >
> > Paul –
> > Thanx and understood.
> > This is a non-normative part of the draft.
> > We are adding advertisement of the remote link identifiers for bundle
> members into the existing L2Bundle Member advertisements specified in RFC
> 8668 (IS-IS) and RFC 9356 (OSPF).
> >
> > But in order to advertise them we have to learn them from somewhere. 😊
> > How that is done is outside the scope of the draft, but we have made a
> couple of suggestions – one of which was to learn them from LLDP.
> > But, as Acee (with your help) has pointed out, we mistakenly referenced the
> Management Address TLV.
> > We will fix that.
> > Acee – I assume that will address your comment? (Thanx for catching this
> mistake)
>
> Yes. You could also add a non-normative reference to
> https://datatracker.ietf.org/doc/draft-ietf-lsvr-l3dl/ as a way to learn the
> remote IDs.
>
> Thanks,
> Acee
>
>
>
>
> > Les
> > From: Paul Congdon
> > <[email protected]<mailto:[email protected]>>
> > Sent: Thursday, March 6, 2025 8:38 PM
> > To: Les Ginsberg (ginsberg)
> > <[email protected]<mailto:[email protected]>>; Acee Lindem
> <[email protected]<mailto:[email protected]>>
> > Cc: [email protected]<mailto:[email protected]>; lsr
> > <[email protected]<mailto:[email protected]>>;
> > draft-glctgp-lsr-l2-<mailto:[email protected]>
> [email protected]<mailto:[email protected]>
> > Subject: Re: WG Adoption Poll for "Advertisement of Remote Interface
> Identifiers for Layer 2 Bundle Members" -draft-glctgp-lsr-l2-bundle-member-
> remote-id-02
> > Les,
> > I have to be honest; I'm not fully up on your draft and I don't know what
> you are trying to accomplish. I was just answering some general LLDP
> questions. To be sure, I'd need to read your draft, but based on the
> exchanges
> I've seen, it seems like a reasonable change. I would suggest that you
> include
> instructions on how to encode the Port ID TLV to meet your specific needs. I
> defer to Acee and others who have read your draft thoroughly.
> > Paul
> > From: Les Ginsberg (ginsberg)
> > <[email protected]<mailto:[email protected]>>
> > Sent: Thursday, March 6, 2025 8:31 PM
> > To: Paul Congdon
> > <[email protected]<mailto:[email protected]>>; Acee Lindem
> <[email protected]<mailto:[email protected]>>
> > Cc: [email protected]<mailto:[email protected]>
> > <[email protected]<mailto:[email protected]>>; lsr
> <[email protected]<mailto:[email protected]>>;
> [email protected]<mailto:[email protected]>
> <draft-<mailto:[email protected]>
> [email protected]<mailto:[email protected]>>
> > Subject: RE: WG Adoption Poll for "Advertisement of Remote Interface
> Identifiers for Layer 2 Bundle Members" -draft-glctgp-lsr-l2-bundle-member-
> remote-id-02
> > Paul –
> > Thanx for the quick response.
> > So if we modify Section 7 of the draft from:
> > “A router may also
> > run LLDP [802.1AB] on the bundle members to exchange local interface
> > identifiers with its neighbor, by using the LLDP Management Address
> > TLV.”
> > To
> > “A router may also
> > run LLDP [802.1AB] on the bundle members to exchange local interface
> > identifiers with its neighbor, by using the Port ID TLV.”
> > Are we on valid ground?
> > Les
> > From: Paul Congdon
> > <[email protected]<mailto:[email protected]>>
> > Sent: Thursday, March 6, 2025 8:26 PM
> > To: Les Ginsberg (ginsberg)
> > <[email protected]<mailto:[email protected]>>; Acee Lindem
> <[email protected]<mailto:[email protected]>>
> > Cc: [email protected]<mailto:[email protected]>; lsr
> > <[email protected]<mailto:[email protected]>>;
> > draft-glctgp-lsr-l2-<mailto:[email protected]>
> [email protected]<mailto:[email protected]>
> > Subject: Re: WG Adoption Poll for "Advertisement of Remote Interface
> Identifiers for Layer 2 Bundle Members" -draft-glctgp-lsr-l2-bundle-member-
> remote-id-02
> > Hello Les,
> > The Port ID TLV is one component of the unique agent identifier. It is
> combined with the Chassis ID TLV to create (what is intended to be) a unique
> identifier within the network scope. It can certainly be used on its own to
> identify the port somehow and you've listed some possibilities.
> > The Management Address TLV was intended to identify an address for
> managing a particular OID, but most commonly used to identify that one
> address used to manage the entire system. If you are looking for a per-port
> address, the Port ID TLV is probably a better choice. In theory, you could
> advertise a bunch of Management Address TLVs and specify different port
> attribute management points (i.e., OIDs), but that would easily fill-up the
> LLDPDU. As a side note, we allow multiple LLDPDUs now-a-days with the
> latest enhancements to LLDP.
> > Paul
> > From: Les Ginsberg (ginsberg)
> > <[email protected]<mailto:[email protected]>>
> > Sent: Thursday, March 6, 2025 8:13 PM
> > To: Paul Congdon
> > <[email protected]<mailto:[email protected]>>; Acee Lindem
> <[email protected]<mailto:[email protected]>>
> > Cc: [email protected]<mailto:[email protected]>
> > <[email protected]<mailto:[email protected]>>; lsr
> <[email protected]<mailto:[email protected]>>;
> [email protected]<mailto:[email protected]>
> <draft-<mailto:[email protected]>
> [email protected]<mailto:[email protected]>>
> > Subject: RE: WG Adoption Poll for "Advertisement of Remote Interface
> Identifiers for Layer 2 Bundle Members" -draft-glctgp-lsr-l2-bundle-member-
> remote-id-02
> > Paul –
> > Thanx for your input – and I certainly will defer to a Layer 2 expert.
> > But I think I confused things by mistakenly referencing the Management
> Address TLV. I should have referenced the Port ID TLV (Section 8.5.3) which
> is a
> Mandatory TLV in every LLDPDU – which has port specific information (not
> Management Address specific information).
> > Apologies for the confusion.
> > As you show below, the Port ID TLV can include:
> > 6 Agent Circuit ID (IETF RFC 3046)
> > If I look at https://datatracker.ietf.org/doc/html/rfc3046#section-3.1
> > Possible uses of this field include:
> > - Router interface number ç===
> > - Switching Hub port number
> > - Remote Access Server port number
> > - Frame Relay DLCI
> > - ATM virtual circuit number
> > - Cable Data virtual circuit number
> > Isn’t the port specific interface number what we have been referring to as
> the “Remote Interface Identifier”?
> > Thanx again for your help.
> > Les
> > From: Paul Congdon
> > <[email protected]<mailto:[email protected]>>
> > Sent: Wednesday, March 5, 2025 11:48 AM
> > To: Acee Lindem <[email protected]<mailto:[email protected]>>
> > Cc: [email protected]<mailto:[email protected]>; lsr
> > <[email protected]<mailto:[email protected]>>;
> > draft-glctgp-lsr-l2-<mailto:[email protected]>
> [email protected]<mailto:[email protected]>;
> Les Ginsberg (ginsberg)
> <[email protected]<mailto:[email protected]>>
> > Subject: RE: WG Adoption Poll for "Advertisement of Remote Interface
> Identifiers for Layer 2 Bundle Members" -draft-glctgp-lsr-l2-bundle-member-
> remote-id-02
> > The Port ID TLV is a mandatory to include TLV in all LLDPDUs and helps form
> the unique identifier for the LLDP agent. There are several ways to represent
> the Port ID.
> > <image001.png> -----Original Message-----
> > From: Acee Lindem <[email protected]<mailto:[email protected]>>
> > Sent: Wednesday, March 5, 2025 11:31 AM
> > To: Paul Congdon <[email protected]<mailto:[email protected]>>
> > Cc: [email protected]<mailto:[email protected]>; lsr
> > <[email protected]<mailto:[email protected]>>;
> > draft-glctgp-lsr-l2-<mailto:[email protected]>
> [email protected]<mailto:[email protected]>;
> Les Ginsberg (ginsberg)
> <[email protected]<mailto:[email protected]>>
> > Subject: Re: WG Adoption Poll for "Advertisement of Remote Interface
> Identifiers for Layer 2 Bundle Members" -draft-glctgp-lsr-l2-bundle-member-
> remote-id-02
> > Hi Paul,
> > That's what I thought - that the management address was not necessarily
> associated with the local L2 interface.
> > The Port ID TLV also doesn't include the local ifIndex so it appears that
> > there
> is currently no way to learn this in LLDP - correct?
> > Thanks,
> > Acee
> > > On Mar 5, 2025, at 1:39 PM, Paul Congdon
> <[email protected]<mailto:[email protected]>> wrote:
> > >
> > > Hello Acee,
> > > Good to hear from you. I've copied Scott Mansfield who worked on the
> LLDP YANG in case he has a different perspective.
> > > The Management Address TLV was designed with a bit of flexibility to
> allow you to advertise a management address used to reach a higher level
> entity for various management objects - not just the local ifIndex of the
> Layer-
> 2 link. I'm not sure anyone has used this flexibility, or if the
> flexibility is
> sufficient for the original intent. Here is some text from the spec:
> > > The Management Address TLV identifies an address associated with the
> > > local LLDP agent that may be used to reach higher layer entities to
> > > assist discovery by network management. The TLV also provides room for
> > > the inclusion of both the system interface number and an object identifier
> (OID) that are associated with this management address, if either or both are
> known.
> > > Here are the usage rules from the spec as well:
> > > 8.5.9.9 Management Address TLV usage rules Management Address TLVs
> are subject to the following:
> > > • At least one Management Address TLV should be included in every
> LLDPDU.
> > > • Since there are typically a number of different addresses associated
> with a MSAP identifier, an individual LLDPDU may contain more than one
> Management Address TLV.
> > > • When Management Address TLV(s) are included in an LLDPDU, the
> included address(es) should be the address(es) offering the best management
> capability.
> > > • If more than one Management Address TLV is included in an LLDPDU,
> each management address shall be different from the management address in
> any other management address TLV in the LLDPDU.
> > > • If an OID is included in the TLV, it shall be reachable by the
> > > management
> address.
> > > • In a properly formed Management Address TLV, the TLV information
> string length is equal to: (management address string length) + (OID string
> length) + 7. If the TLV information string length in a received Management
> Address TLV is incorrect, then it is ignored and processing of that LLDPDU is
> terminated.
> > > Hope this is helpful,
> > > Paul
> > > -----Original Message-----
> > > From: Acee Lindem <[email protected]<mailto:[email protected]>>
> > > Sent: Wednesday, March 5, 2025 9:42 AM
> > > To: [email protected]<mailto:[email protected]>
> > > Cc: lsr <[email protected]<mailto:[email protected]>>;
> > > [email protected]<mailto:[email protected]>;
> > > Les Ginsberg
> > > (ginsberg) <[email protected]<mailto:[email protected]>>
> > > Subject: Re: WG Adoption Poll for "Advertisement of Remote Interface
> > > Identifiers for Layer 2 Bundle Members"
> > > -draft-glctgp-lsr-l2-bundle-member-remote-id-02
> > > Hey Paul, Is the interface number associated with the LLDP Management
> Address TLV always the local ifIndex of the Layer-2 link?
> > > Hope All is Well,
> > > Acee
> > > > On Mar 4, 2025, at 2:01 PM, Les Ginsberg (ginsberg)
> <[email protected]<mailto:[email protected]>> wrote:
> > > > > Acee -
> > > > IEEE Std 802.1AB-2016 Figure 8-11 has exactly what we need.
> > > > In particular:
> > > > 8.5.9.5 interface numbering subtype
> > > > The interface numbering subtype field shall contain an integer value
> > > > indicating the numbering method used for defining the interface number.
> The following three values are currently defined:
> > > > 1) Unknown
> > > > 2) ifIndex
> > > > 3) system port number
> > > > And
> > > > 8.5.9.6 interface number
> > > > The interface number field shall contain the assigned number within
> > > > the system that identifies the specific interface associated with this
> management address.
> > > > Les
> > > >> -----Original Message-----
> > > >> From: Acee Lindem <[email protected]<mailto:[email protected]>>
> > > >> Sent: Tuesday, March 4, 2025 9:56 AM
> > > >> To: Acee Lindem <[email protected]<mailto:[email protected]>>
> > > >> Cc: lsr <[email protected]<mailto:[email protected]>>;
> > > >> [email protected]<mailto:[email protected]>
> > > >> Subject: Re: WG Adoption Poll for "Advertisement of Remote
> > > >> Interface Identifiers for Layer 2 Bundle Members" >>
> > > >> -draft-glctgp-lsr-l2-bundle-member-
> > > >> remote-id-02
> > > >>> Speaking as WG member:
> > > >>> I support adoption.
> > > >>> With respect to acquiring the remote ID, I don't believe that LLDP
> > > >>> include the
> > > >> remote ID. There is a port ID but I believe this is an L2 construct.
> > > >> If you're going to reference LLDP, you should add it to the "LLDP
> > > >> IETF Organizationally Specific TLV" as is done for BGP parameters
> > > >> in
> > > >> https://da/
> > > >>
> ta%2F&data=05%7C02%7C%7C2ef878cdb8384b8874e008dd5c1c4a84%7C
> 84df9e7f
> > > >>
> e9f640afb435aaaaaaaaaaaa%7C1%7C0%7C638767998757179357%7CUnk
> nown%7CT
> > > >>
> WFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJX
> aW4z
> > > >>
> MiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=ZQy%2B
> vaB4Q
> > > >> luy3OjkcXMxhy1pCDBBiWRzGfkK1kqKKcw%3D&reserved=0
> > > >> tracker.ietf.org%2Fdoc%2Fdraft-
> &data=05%7C02%7C%7C0fdffaebc088410bc
> > > >> b3
> > > >>
> 808dd5c0d0812%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C63
> 876793
> > > >> 32
> > > >>
> 21668063%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIl
> YiOiIwLj
> > > >> Au
> > > >>
> MDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C
> %7C%
> > > >> 7C
> > > >>
> &sdata=qBFUmw7pepHx3vPAxxxaCs3rfQi1RgKj6wgwRw2GJjY%3D&reserved
> =0
> > > >> acee-idr-lldp-peer-discovery/. Also, you don't mention (via an
> > > >> informational
> > > >> reference) >> https://datatracker.ietf.org/doc/draft-ietf-lsvr-l3dl/
> > > >> which
> does, in fact, advertise the local ifIndex which is commonly used as the
> interface ID.
> > > >>> Thanks,
> > > >> Acee
> > > >>> On Mar 2, 2025, at 8:56 AM, Acee Lindem
> > > >>> <[email protected]<mailto:[email protected]>>
> wrote:
> > > >>> >>> LSR WG,
> > > >>> >>> This starts the Working Group adoption call for >>>
> > > >>> >>> draft-glctgp-lsr-l2-bundle-
> > > >> member-remote-id-02. Please send your
> > > >>> support or objection to this list before March 17th, 2025.
> > > >>> >>> Thanks,
> > > >>> Acee
> > > >>
>
_______________________________________________
Lsr mailing list -- [email protected]
To unsubscribe send an email to [email protected]
