[ih] History question -- multi-homing

[Brian E Carpenter]

>>> what happens if two corresponding nodes change their locators at the same
>>> time?
There is considerable discussion of this issue in RFC5533 (SHIM6, one of the
unsuccessful attempts to fix this issue for IPv6). There are two states each
end can encounter if both ends need to change locators at the same time
("context loss" and "context confusion"). In the worst case, of course, the
only solution is the well known "switch it off and on again" method.

I think this is probably a theorem - there are always going to be residual
race conditions in which a session will fail.

If Lixia was here, she'd be reminding us about Named Data Networking
(https://named-data.net/).

Regards/Ngā mihi
    Brian

On 13-Sep-26 22:55, vinton cerf via Internet-history wrote:
> seems as if they might lose each other unless there is confirmation of the
> receipt of a new locator before the change? Either that or some agreed
> rendezvous point that doesn't change?
> 
> Do you know how this is handled in mobile telephony? Is there a moment when
> multiple base stations are aware of or serve the same mobile call? Of
> course, the base stations don't "move" so that helps, since the mobiles
> know about signal levels from multiple base stations.
> v
> 
> 
> On Sun, Sep 13, 2026 at 6:47 AM John Day <jeanjour at comcast.net> wrote:
> 
>> Depends on whether their point of attachment changes. Each is somewhere at
>> any point in time.
>>
>> Or as Firesign Theater said:
>> How Can You Be in Two Places at Once, When You Are No Place At All.  ;-)
>>
>>> On Sep 13, 2026, at 05:19, vinton cerf via Internet-history <
>> internet-history at elists.isoc.org> wrote:
>>>
>>> what happens if two corresponding nodes change their locators at the same
>>> time?
>>>
>>> v
>>>
>>>
>>> On Sat, Sep 12, 2026 at 12:15 PM vinton cerf <vgcerf at gmail.com> wrote:
>>>
>>>> thanks tony.
>>>> v
>>>>
>>>>
>>>> On Sat, Sep 12, 2026 at 10:49 AM Tony Li <tony.li at tony.li> wrote:
>>>>
>>>>>
>>>>> Hi Vint,
>>>>>
>>>>>> It would be an interesting exercise to posit a different
>> identification
>>>>>> scheme that would scale and also permit multi-homing. Has someone
>>>>> already
>>>>>> worked that out? Would it require a flat ID space?
>>>>>
>>>>>
>>>>> Please see RFC 6740.
>>>>>
>>>>> Tony
>>>>>
>>>>>
>>> --
>>> Internet-history mailing list
>>> Internet-history at elists.isoc.org
>>> https://elists.isoc.org/mailman/listinfo/internet-history
>>> -
>>> Unsubscribe:
>> https://app.smartsheet.com/b/form/9b6ef0621638436ab0a9b23cb0668b0b?The%20list%20to%20be%20unsubscribed%20from=Internet-history
>>
>>