The /128 issue with Linode (insofar as it relates to Spamhaus) has been
percolating for at least 5 years if I recall correctly, but no shorter
than the strong RFC-level guidelines of /64.
Linode will allocate at /64 on request, and has being doing so for about
just as long.
On 2021-11-25 2:01 p.m., Jarland Donnell via mailop wrote:
In all fairness, some of these systems may have been deployed before we
were all really certain that a /64 per customer was going to be an
accepted standard. You know how RFCs go, they're the law of the land
except when they're not, which is actually pretty often. By now most
should have figured out that they need to conform to this one, but I
don't really blame early adopters for second guessing what common
implementations would look like a few years down the road.
On 2021-11-25 12:04, Jay Hennigan via mailop wrote:
On Thu, Nov 25, 2021 at 03:07:02PM +0200, Mary via mailop wrote:>> I
think Linode does not follow the /64 rule and assigns thousands of
customers within the 2a01:7e01::/64 block. They user a bunch of
blocks, depending on their data centre.>> I think by default each
client is assigned a single /128 IPv6 address per server.
That is rather stupid behavior on Linode's part then. The rest of the
Internet uses a /64 per subnet and typically a /56 per customer
minimum.
What are they thinking? Are they really worried about running out of
IPv6 addresses?
Vote with your feet.
_______________________________________________
mailop mailing list
[email protected]
https://list.mailop.org/listinfo/mailop
_______________________________________________
mailop mailing list
[email protected]
https://list.mailop.org/listinfo/mailop
