Hi Acee, Thank you for the update!
Em seg., 6 de out. de 2025 às 09:06, Acee Lindem <[email protected]> escreveu: > > 2 - The "msd-type" leaves use an identityref type whose base is > > "iana-msd-types:msd-base-mpls". I see two issues here: first, the MSD > > signaling specified by RFC 8491 isn't restricted to MPLS MSDs. Second, by > > using an identityref, unknown MSD types cannot be represented. For this > > reason, I think using a uint8 for the "msd-type" leaves would be better, as > > was done in the OSPF module. > > Actually, the OSPF MSD model had both the identityref and the unit8. I added > the raw value to the grouping. Hmm, I wasn't aware there was an OSPF MSD module as well. I was referring to the MSD nodes present in the main OSPF module, specifically this grouping: https://github.com/YangModels/yang/blob/main/standard/ietf/RFC/ietf-ospf%402022-10-19.yang#L1320 I checked the OSPF MSD module and it seems the OSPFv2 augmentations are unnecessary, as the base module already defines MSD nodes for OSPFv2. You can see it here: +--ro ospf:opaque +--ro ospf:ri-opaque | +--ro ospf:maximum-sid-depth-tlv | | +--ro ospf:msd-type* | | +--ro ospf:msd-type? uint8 | | +--ro ospf:msd-value? uint8 | +--ro ospf-msd:node-msd-tlv | +--ro ospf-msd:node-msds* [msd-type] | +--ro ospf-msd:msd-type identityref | +--ro ospf-msd:msd-value? uint8 This is only for OSPFv2. For OSPFv3, the "ietf-ospfv3-extended-lsa" module doesn't define any nodes for the MSD TLVs. Back to the original issue, if we want to have both the identityref and the raw value, shouldn't the identityref use "msd-base" instead of "msd-base-mpls" as its base? I see "ietf-ospfv3-srv6" defines an "msd" container with four leafs for each SRv6 MSD type. Is there a reason for not having a single MSD list that covers all possible MSD types, including both MPLS and SRv6? Best Regards, -- Renato Westphal _______________________________________________ Lsr mailing list -- [email protected] To unsubscribe send an email to [email protected]
