Re: [Autoconf] Call for comments to a new AUTOCONF charter proposal.

[Alexandru Petrescu <[e-mail]>]

Innserting some comments although I'm not addressed.

Le 30/06/2010 00:16, Ryuji Wakikawa a écrit :
> Hi Henning and all,
>
> On 2010/06/29, at 5:44, Henning Rogge wrote:
>
>> On Tue June 29 2010 14:34:41 Alexandru Petrescu wrote:
>>>> I don't think elections are a good way to add redundancy for an
>>>> address configuration system. The dependency of a single node
>>>> in the whole network to configure the addresses will be a
>>>> bottleneck for larger mesh networks.
>>>
>>> But elections is what gets rid of central node bottlenecks... if
>>>  that single node fails then elect another.
>> Elections are a real complex thing in MANETs, better have
>> redundancy and "no bottlenecks" by allowing multiple redundant
>> address providers.
>
>> From the proposed charter, election part is just "future extension"
>> and not the item we will work on right away.
>
> In this charter we try to recycle existing IP autoconfiguration
> protocols for MANET environment.
>
> This was clearly noted in the previous charter. see below. "This
> group's effort may include the development of new protocol
> mechanisms, should the existing IP autoconfiguration mechanisms be
> found inadequate. However, the first task of the working group is to
>  describe one practical addressing model for ad hoc networks."
>
> DHCP is a protocol supporting prefix delegation and address
> assignment now. RA based approach (Alex's proposal) may be an option,
> too. However, 6LowPAN WG is working on this subject based on the
> AUTOCONF address arch.

Well, an RA-based prefix exchange and route update would not be
dedicated to 802.15.4 only (6LoWPAN), and not to PAN either.  I run it
on wifi between Mobile Routers, supposedly between vehicles, and not on
a PAN.

> Why not stick to the existing IP autoconfiguration mechanism as a
> starting point so that we can directly go to the solution space
> discussion. Do you want to spend a few years to define protocol
> requirements of a solution(s)? I specially concern the endless
> activity cycle in AUTOCONF from the WG's history.

Hmmm... I tend to agree to stick with existing IP address
autoconfiguration mechanisms in AUTOCONF.

>>>> I don't see a difference between 1 and 2, for IPv6 case 1 is
>>>> just case 2 with a prefix length of 128 (or 64 if we use the
>>>> IPv6 autoconf postfix).
>>>
>>> To my quick reading 2) does prefixes too whereas 1) does
>>> addresses only.
>> Which make it a waste of time to develop sollution 1, because it's
>>  only a special case of sollution 2.
>
> Why waste of time? These two items should be tightly related, but
> each solution (1 and ) should work independently. For instance, we
> may use a manet routing protocol to assign an address and use DHCP
> for PD in the future.

(I set aside what you are saying, explicitly the part suggesting that
maybe the MANET routing protocol may assign an address... if so then, by
symmetry, the address autoconfiguration mechanism could also exchange
routing information and update the routing information databases -
necessary for DHCP sometimes).

> To support these future extension, we should separate these two
> items.

I agree I think we should follow the existing separation of prefixes vs
addresses when it comes to address autoconfiguration mechanisms DHCPv6
and DHCPv6 PD.

I need a new term like "address and prefix autoconfiguration mechanism",
something like SLPAC (prefix autoconf instead of SLAAC), something like
DRCP (Dynamic Router Configuration Protocol instead of DHCP).

Alex

>
> regards, ryuji
>
>
>>
>
>>
>> Henning Rogge -- Diplom-Informatiker Henning Rogge ,
>> Fraunhofer-Institut für Kommunikation, Informationsverarbeitung und
>> Ergonomie FKIE Kommunikationssysteme (KOM) Neuenahrer Straße 20,
>> 53343 Wachtberg, Germany Telefon [phone],   Fax +49 228
>> 9435 685 mailto:[e-mail]
>> http://www.fkie.fraunhofer.de GPG: E1C6 0914 490B 3909 D944 F80D
>> 4487 C67C 55EC CFE0
>> _______________________________________________ Autoconf mailing
>> list [e-mail]
>> https://www.ietf.org/mailman/listinfo/autoconf
>
> _______________________________________________ Autoconf mailing
> list [e-mail]
> https://www.ietf.org/mailman/listinfo/autoconf
>