Alvaro, Yes, your suggested text would be fine with me.
Thanks, Donald =============================== Donald E. Eastlake 3rd +1-508-333-2270 (cell) 2386 Panoramic Circle, Apopka, FL 32703 USA [email protected] On Wed, Jun 1, 2022 at 10:31 AM Alvaro Retana <[email protected]> wrote: > On June 1, 2022 at 10:06:58 AM, Luigi Iannone wrote: > > > > On 1 Jun 2022, at 15:46, Luigi Iannone wrote: > > > > > > Hi Alvaro, > > > > > >> On 1 Jun 2022, at 15:29, Alvaro Retana wrote: > > >> Nice sentence, but what does it mean? What action should the > > >> implementation take to satisfy "MUST be managed"? From Luigi's > > >> initial response (above), it sounds like the time between changes (TTL > > >> ?) has to be long enough... Please spell some of this out. > > >> > > > > > > What if we change to: > > > > > > Mapping updates, and their corresponding Map Version Number MUST be > managed > > > so that a very old version number will not be confused as a new version > > > number (because of the circular numbering space). > > > > > > > Or may be: > > > > Mapping updates, and their corresponding Map Version Number MUST be > managed > > so that a very old version number will not be confused as a new version > > number (because of the circular numbering space). To this end simple > measures > > can be taken, like Updating a mapping only when all active traffic is > using > > the latest version, or waiting sufficient time to be sure that mapping in > > LISP caches expire, which means waiting at least as much as the mapping > Time- > > To-Live (as defined in 6833bis) > > > > Any preference? > > The normative text is what is not clear enough for me: "MUST be > managed" is not actionable. The examples help. I would be fine if > the normative language is removed while still mentioning that it is a > requirement: > > Mapping updates, and their corresponding Map Version Numbers have to be > managed so that a very old version number will not be confused as a new > version number (because of the circular numbering space). To this end... > > > Donald: is this ok with you? > > > Thanks! > > Alvaro. >
_______________________________________________ lisp mailing list [email protected] https://www.ietf.org/mailman/listinfo/lisp
