Joel: Before implementation, we must assure it is feasible. Obviously, the current proposal is impractical.
As Robert has pointed also out, the length limit for the every potential MP-TLV is sky(unlimited length) and forever(the receiver doesn’t know when to stop parsing to form the original big-TLV) Aijun Wang China Telecom > On Mar 29, 2025, at 09:19, Joel Halpern <j...@joelhalpern.com> wrote: > > That sounds like it rather depends upon your implementation approach? > > Yours, > > Joel > >> On 3/28/2025 9:07 PM, Aijun Wang wrote: >> There is no necessary to concatenate the difference type IS-IS TLVs >> together, but it must be done for the MP-TLV pieces. >> >> I have never seen any example that there is no any length signaling for >> segmentation one big-TLV and can concatenate it correctly without the right >> length indication. >> >> It’s pity the LSR WG will endorse such proposal. >> >> >> >> Aijun Wang >> China Telecom >> >>>> On Mar 29, 2025, at 08:16, Tony Li <tony...@tony.li> wrote: >>> >>> >>> Hi Robert, >>> >>>> Sorry if I was not very clear but my point was that MP-TLV may be coming >>>> in multiple LSPs - something which to the best of my understanding is not >>>> the case today with any TLV type. On that basis as LSP length natural >>>> boundary is gone it seems sky is the limit now. >>> >>> This is incorrect. A TLV may appear in many different framgments. For >>> example, a router may have multiple adjacencies. These may be spread out >>> across fragments. >>> >>> T >>> >>> >>> _______________________________________________ >>> Lsr mailing list -- lsr@ietf.org >>> To unsubscribe send an email to lsr-le...@ietf.org >> _______________________________________________ >> Lsr mailing list -- lsr@ietf.org >> To unsubscribe send an email to lsr-le...@ietf.org > _______________________________________________ Lsr mailing list -- lsr@ietf.org To unsubscribe send an email to lsr-le...@ietf.org
