On Wed, 16 Oct 2024, Randy Bush wrote:
Name: draft-gasser-opsawg-prefix-lengths
URL:
https://www.ietf.org/archive/id/draft-gasser-opsawg-prefix-lengths-01.txt
Also take a look at my draft draft-levine-6man-repsize which we have
discussed in 6man. It proposes a design that is essentially identical.
no. aside from lightness of motivation, it does not cover ipv4. this
is the OPS area, and we operators are still stuck with ipv4 being
dominant.
Indeed it is, but I don't see a lot of networks doing IPv4 suballocations
on power-of-two boundaries these days. I don't think there's anything
wrong with letting people publish v4 prefixes but I also don't think it's
very useful.
For example, if you get 4 or 8 hosts at a provider like AWS or Digital
Ocean, you will get random IPs from their pool, not a /30 or a /29.
On my own physical server, I have legacy /24 that belongs to one of my
users from which I've let the local network use some /30's at the low end
for their other customers, while the high 192 addresses are mine.
On my home broadband, I have an individual address but some of the other
customers are behind a CGNAT.
I suppose you could invent a much more complex language to describe all
these different shapes but I don't see the point, since it's hard to
imagine legit networks publishing that level of detail. Sleazy networks
will publish whatever lies they think will get their spam delivered but
there's nothing new about that, they've been "firing bad customers" and
pretending they don't control those IPs for years.
there is a bunch of micro stuff we learned in the 9092/9632 process.
e.g. the remarks: token, Prefixlen is capitalized because RPSL folk and
rfced.
Yes, those are better. Thanks.
at least draft-gasser-opsawg-prefix-lengths acknowledges that it is a
riff off 9092/9632 and its predecessors by much of the same author list.
[ we're chasing 8805 folk. ] giving cred is painless and helps form a
positive culture.
I don't understand this issue. My draft refers to 9632 and says to sign
the prefix files the same way that 9632 signs geoip files. That seemed
better than copying all the text and hoping I got it right.
I haven't looked at your draft in enough detail to see whether your
signatures are the same as 9632 or slightly different. If they're
different it would be helpful to call out the differences since I expect
people will use the same signing routines and will need to know what to
change.
R's,
John
_______________________________________________
OPSAWG mailing list -- opsawg@ietf.org
To unsubscribe send an email to opsawg-le...@ietf.org
