On 07/02/2022 16:03, Templin (US), Fred L wrote:
I will repeat what I said; getting this right the first time is trivial and ignoring it the first time ruins the opportunity to use it for bigger and better purposes for forever. Bob objects that there has been little uptake in jumbograms, but the correct statement is that there has been little uptake *until now* - IP Parcels will change all of that. And, not just for the local link, but over the entire path. An example: Source -> 4M -> 2M -> 512K -> 256K -> 1M -> 3M -> Destination. In that case, the MTU option had better report 256K. I will now briefly take up the subject of bull-throwing. I think many of the people we have seen in these discussions and elsewhere are very skilled in the art of throwing the bull. Some virtuous few others are genuine in their approach and want the best technical outcome. I would say to them be authentic; be genuine; be yourself, and do not succumb to the ways of the hardcore bull-throwers. Thanks - Fred
I'm sorry but I also agree with Bob: I don't see that your proposal relates to the IPv6 deployment scenario in the IETF-LC. At this time, we haven't seen vendors with support for more than 16 KB frames (for various reasons I'll let them explain), and despite some deployment in DCs we have yet to see many operational Internet paths that allow more than ~ 2 KB.
I'll look forward with interest to seeing the level of support to deploy 4MB MTUs (or whatever size you target) in INTAREA and how challenges will be addressed, but I don't see this as a great reason for growing the size of the IPv6 HBH PMTU option.
Gorry -- last-call mailing list last-call@xxxxxxxx https://www.ietf.org/mailman/listinfo/last-call
