Re: [16NG] Resolutions to issues raised during 2nd WGLC of I-D draft-ietf-16ng-ipv6-over-ipv6cs-04 [1]

Hi Syam, thanks for the reply,

Syam Madanapalli wrote:
> some text inline.
> 
>> 
>> JinHyeock Choi wrote:
>>> Dear Alex
>>> 
>>> Thanks for your feedback. Kindly find my in-line comments.
>>> 
>>>>> and there is no reason for the host to depend on RAs to 
>>>>> detect if it has moved across ARs. Hence there is no reason 
>>>>> for any change w.r.t 3775 and I do not see RFC3775 
>>>>> over-riding the values specified here.
>>>> 
>>>> There is no other means for an IP stack on top of IPv6CS to 
>>>> know that the subnet has changed.  This is where MIP6 reacts, 
>>>> on subnet change, not on PHY change.  So the 802.16 BS, if it 
>>>> sends RAs, it should send them quickly.
>>> 
>>> Allow me to make is more clear.
>>> 
>>> A 802.16 host doesn't have to rely on unsolicited PERIODIC RAs 
>>> for movement detection. Upon network attachment, either 1) the 
>>> access router sends an unsolicted RA as of FRD 
>>> (draft-ietf-dna-frd-02) or 2) the host sends an RS to get a 
>>> solicited RA as of DNAv6 (draft-ietf-dna-protocol-03.txt).
>> 
>> I am not sure what you mean by 'network attachment'?  I suppose you
>>  mean the network entry procedure of 802.16?  ('Attach' in 802.16 
>> mostly means to attach a part of message to another part, and only
>>  in one single place a SS to a BS). In the network entry procedure
>>  of the 802.16 document there is nothing that suggests the
>> behaviour you suggest above (DNA).   Only ND is suggested, not DNA.
>> 
> 
> DNA is nothing to do with 802.16, it is layer 3 mechanism, you can 
> find the details at 
> http://www.ietf.org/internet-drafts/draft-ietf-dna-protocol-03.txt

Yes, it is a layer 3 mechanism.  Only layer 3 mechanisms can be designed
at IETF, or layer-2-3, but not layer 3.  I think we agree.

>> Periodic RA is in widespread use for movement detection.  I do not 
>> understand the opposition to use periodic RAs as allowed by 
>> rfc3775. What breaks if we use periodic RAs with a frequency higher
>>  than 1/4s? (remark arguments of air interface gains are different
>>  on high-bandwidth links).
> 
> Frequent periodic RAs consume the air bandwidth and may cause 
> problems for dormant devices.

I agree they may be an issue.  But to what extent?

Dormant devices always have something listening and executing code.

> Even if the above is not a problem, still depending RAs for movement
>  detection may not be a good idea. The better way is to solicit for
> an RA when the device observes a L2 movement, to check if the device
> is on the same link or new one.

YEs, it is a very good idea to send an RS to get an RA and compare the
prefix see whether MIP6 needs to be triggered.  But where to send this
RS?  To the IP all-nodes multicast address?  I agree it works to send it
and the BS will receive it, and respond RA.  But remark it will receive
it even if SS puts the digit '5' there, not necessarily 'ff02::1'.
Which to me is a door open to incompatibilities and potential
inter-operability issues.

But, if it's thought that this is not an issue I will not insist.  I
agree with the group.  The mechanism to send an RS is good enough to
detect a subnet-layer handover and many implementations can do.

Alex

_______________________________________________
16NG mailing list
16NG@ietf.org
https://www1.ietf.org/mailman/listinfo/16ng