[pim] draft-ietf-pim-zeroconf-mcast-addr-alloc-ps issue(s)
Toerless Eckert <[email protected]> Tue, 23 Dec 2025 00:26:21 +0100
| Newsgroups | gmane.ietf.pim |
|---|---|
| Message-ID | <[email protected]> |
Just doing RTGDIR review of draft-ietf-pim-ipv6-zeroconf-assignment
and stumbling across the following in subject draft:
So far, the whole justification of the draft, avoidance of collisions
at link-layer is only written out in the explanatory text, but not
mentioned in the requirements:
[REQ-1] Unique Address Assignment: Use of the solution SHALL
result in a unique address being assigned to the multicast group.
replace for example:
[REQ-1] Unique Address Assignment: Use of the solution SHALL
result in an assignment of an IPv4 and/or IPv6 multicast address
that is unique, and whose link-layer address is also unique.
also:
[REQ-8] Collision Detection and Resolution: The solution SHALL
include mechanisms to detect and resolve multicast address
collisions.
maybe:
[REQ-8] Collision Detection and Resolution: The solution SHALL
include mechanisms to detect and resolve multicast address
collisions including collisions at the link-layer.
Haven't thought about the right wording in too much detail. But
its IMHO front and center to ensure the link layer collision
avoidance is explicitly mentioned in those requirements.
more minor:
I can not find an explanation as to what CONS-6 would be good for.
Some example would be good. Is this for example about enumeration
of all used addresses for network management purposes ? Then maybe
rephrase accordingly "Third-party (network management) application
SHOULD be capable to determine all currently used addresses".
CONS-7 is also unclear. Again, think of explicit goal and try to
rephrase accordingly.
Cheers
Toerless
_______________________________________________
pim mailing list -- [email protected]
To unsubscribe send an email to [email protected]