Re: [Idr] AS Number in TLV 512 (rfc7752bis / draft-ietf-lsvr-bgp-spf)

[Ketan Talaulikar <[e-mail]>]

Hi Alvaro,

Please check inline below.

On Fri, Mar 10, 2023 at 8:15 PM Alvaro Retana <[e-mail]>
wrote:

> Ketan:
>
> Hi!
>
> While reviewing draft-ietf-lsvr-bgp-spf I came across this definition
> of AS for TLV 512:
>
>    Autonomous System:  Opaque value (32-bit AS Number).  This is an
>    optional TLV.  The value SHOULD be set to the AS Number associated
>    with the BGP process originating the link-state information.  An
>    implementation MAY provide a configuration option on the BGP-LS
>    Producer to use a different value; e.g., to avoid collisions when
>    using private AS numbers.
>
>
> The text doesn't require the ASN to be the same as what is on the BGP
> OPEN.  It even let's any ASN be used, which could result in other
> collisions or confusion. :-(
>

KT> It SHOULD be the ASN of the BGP-LS Originator but that is not a MUST.
Consider that an operator does not wish to use their "real" ASN for BGP
sessions that are only being used for BGP-LS reporting. But they may want
to use the "real" one in the NLRI in some scenarios when doing multi-AS
topology consolidation. That is why some implementations allow for the
operator to configure/provide a different ASN to be used within the BGP-LS
NLRI.


> This definition was in rfc7752, and I don't remember what the
> justification was for it, do you?
>
> It seems to me that using the BGP-LS Instance-ID as a differentiator
> of the origin domain would solve any issues related to multiple
> autonomous systems with the same ASN -- in line with the
> recommendation for IGPs.


KT> The draft does in fact suggest that in section 3.2.14:

   The BGP-LS Instance-ID carried in the Identifier field as described
   earlier along with a set of sub-TLVs described in Section 5.2.1.4
<https://datatracker.ietf.org/doc/html/draft-ietf-idr-rfc7752bis#section-5.2.1.4>,
   allows specification of a flexible key for any given node/link
   information such that the global uniqueness of the NLRI is ensured.
   Since the BGP-LS Instance-ID is operator assigned, its allocation
   scheme can ensure that each IGP domain is uniquely identified even
   across a multi-AS network.




> It would also allow the definition above to
> say that the "value SHOULD be set to the AS Number associated with the
> BGP process originating...".
>

KT> However, we have existing deployments and asking for a
renumbering/coordination of BGP-LS Instance ID across a multi-AS
multi-IGP-domain network might be operationally challenging. Therefore, we
still have the ASN so it is possible for the BGP-LS Instance-ID (that is
associated with an IGP domain) to be unique within that AS.

Thanks,
Ketan


>
>
> Thoughts?
>
> Thanks!
>
> Alvaro.
>