Re: [Lsr] [draft-ietf-lsr-flex-algo-08] Clarification on ASLA usage for flex-algo

[Peter Psenak <[e-mail]>]

Hi Gunter,

On 06/08/2020 18:31, Van De Velde, Gunter (Nokia - BE/Antwerp) wrote:
> Hi Authors, All,
> 
> My understanding is that for new LSR applications we should select 
> either “ASLA encoding” or select “legacy encoding” for all Link attributes.
> 
> Not a mixture of both. There is a clear long term technology benefit of 
> using all ASLA encoding.
> 
> In draft-ietf-lsr-flex-algo-08 “Section 11: Advertisement of Link 
> Attributes for Flex-Algorithm” we find that use of ASLA is mandated
> 
> “
> 
>    Link attribute advertisements that are to be used during Flex-
> 
>     Algorithm calculation MUST use the Application Specific Link
> 
>     Attribute (ASLA) advertisements defined in [I-D.ietf-isis-te-app] or
> 
>     [I-D.ietf-ospf-te-link-attr-reuse].
> 
> “
> 
> This is a ‘normative’ instruction that link attributes MUST use ASLA 
> encoding.
> 
> Hence link attributes like (e)ag, te-metric and delay MUST be ASLA encoded.
> 
> Is that correct understanding?

yes, for any link attributes used for flex-algo.

> 
> If yes, it is odd that some attributes are new ASLA and some other are 
> non-future proof ‘LEGACY’ encodings
> 
>   * Example#1: section 5.1 “ISIS Flexible Algorithm Definition Sub-TLV”
>     and section 5.2 “OSPF Flexible Algorithm Definition TLV” point
>     towards legacy encodings RFC7810 and RFC5305 (idnit RFC7810 is
>     obsoleted by RFC8570).
>   * Example#2: section 5.2 “OSPF Flexible Algorithm Definition TLV” the
>     normative references are missing

ASLA is link-attributes related (Application Specific Link Attributes).

ASLA can only be used to encode link attributes. FAD is not a link 
attribute, so it can not use ASLA.

thanks,
Peter

> 
> Sections 5.1 and 5.2 should update the link attribute encoding 
> references to reflect ASLA encoding as specified in section 11.
> 
> The text referencing legacy encodings should be removed
> 
> G/
>