Re: [v6ops] A common problem with SLAAC in "renumbering" scenarios

[Fernando Gont <[e-mail]>]

On 31/1/19 10:51, Richard Patterson wrote:
> On Thu, 31 Jan 2019 at 12:15, Fernando Gont <[e-mail]> wrote:
>>
>> On 31/1/19 08:40, Tassos Chatzithomaoglou wrote:
>>> Isn't that handled by L-13 of RFC7084?
>>>
>>> L-13:  If the delegated prefix changes, i.e., the current prefix is
>>>           replaced with a new prefix without any overlapping time
>>>           period, then the IPv6 CE router MUST immediately advertise the
>>>           old prefix with a Preferred Lifetime of zero and a Valid
>>>           Lifetime of either a) zero or b) the lower of the current
>>>           Valid Lifetime and two hours (which must be decremented in
>>>           real time) in a Router Advertisement message as described in
>>>           Section 5.5.3, (e) of [RFC4862].
>>
>> Yes, we missed this REQ, and will update our doc accordingly. HOwever,
>> this bullet was included for completeness sake. The case we care and are
>> trying to address here is that in which the CPE reboots.
> 
> I think this requirement already resolves the CPE-side of things, and
> the rest I believe, is solved server-side?
> If the CPE is hard rebooted, it hasn't sent a Release, so I'd expect
> the lease-state to be kept server-side, and re-offered if the DUID+MAC
> binding is the same when the CPE reconnects.

Well, if you want to play safe, you cant expect anything. Also, in some
cases it may be the actual *goal* to get a *different* prefix. i.e., the
so-called "dynamic" prefixes, which seem to be the recommended default
there.



> Although, during soft reboots, I have often seen CPEs sending a
> Release, telling the servers to relinquish the lease.  Perhaps we
> should discuss the benefits/detriments of doing this?

I will add some text on the topic. However, if we elaborate more on the
CPE behaviour we may want to end up splitting the document: one part
that focuses on the host-behaviour and that is aimed at 6man, and the
CPE part aimed at v6ops?

Thanks!

Cheers,
-- 
Fernando Gont
SI6 Networks
e-mail: [e-mail]
PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492