Re: [OSPF] Anycast Extension to OSPFv3 draft-wang-aospf-00

[Acee Lindem <[e-mail]>]

Hi Greg,
Greg Mirsky wrote:
> Dear Abhay,
> I apologize for non-technical post below.
> If an operator agrees with running someone's proprietary protocol would
> operator care about IETF's and OSPF WG's seal of approval? But if such
> seal is required then it's up to the group to decide whther to put it
> down or not. Would you agree?
> Any group can use any grant to re-work any existing interoperable
> implementation to make it "different". And probably it will be fun and
> educational experience to the group. Should OSPF WG or IETF as whole be
> involved in such projects?
>   
Dependent on WG bandwidth, we could allow extensions to be documented as
experimental RFCs (let's ignore the statement about a separate version 
number below :^).
However, the WG would still need to agree the to accept the draft as a 
viable
solution to an accepted problem. In this case, my opinion is that this 
draft shouldn't
be accepted by the WG.

Thanks,
Acee

> 	Regards,
> 		Greg
>
> -----Original Message-----
> From: Abhay D.S [mailto:[e-mail]] 
> Sent: Wednesday, December 13, 2006 3:30 AM
> To: Acee Lindem
> Cc: [e-mail]
> Subject: Re: [OSPF] Anycast Extension to OSPFv3 draft-wang-aospf-00
>
> hi acee,
>
> We use OSPF as a TOOL which can be programmed for various styles.
>
> For example:
>
> I want one version of OSPFv3 version running WG approved stamp on it and
>
> inter-operating.
>
> However I can have my own versions of OSPF probably with a different 
> version number within
> my network  "only to process anycasting". Probably because anycasting is
>
> bursty.
>
> Precisely the new AOSPF draft does that and there is nothing wrong with 
> it. Tell me what is
>
> wrong in this requirement ?
>
> Thanks,
> Abhay
>
>
>
>
>
>
>
>
>
>
> Acee Lindem wrote:
>   
>> Abhay,
>>
>> Abhay D.S wrote:
>>     
>>> hi acee,
>>> With due respect to your disposition, the Anycast draft is a NSF 
>>> grant project
>>> and it makes a good implementation, as good as OSPFv3 anycast 
>>> capability.
>>>       
>> On what technical grounds do you based this statement? I don't see any
>>     
>
>   
>> reason
>> to handle anycast addresses any differently from any other /128
>>     
> prefix.
>   
>>> The authors are definitely, trying out to reach people with other 
>>> perspectives which is
>>> probably different from ours.
>>> Hope you take positive criticism positively.
>>>       
>> Well that doesn't change the fact that protocol extensions are not 
>> automatically
>> accepted by the OSPF WG.
>>
>> Acee
>>
>>     
>>> Period.
>>> Thanks,
>>> Abhay
>>>
>>>
>>>
>>>
>>> Acee Lindem wrote:
>>>       
>>>> Abhay,
>>>>
>>>> Abhay D.S wrote:
>>>>         
>>>>> hi acee,
>>>>> Not very positive and encouraging, being on a WG chair.
>>>>>           
>>>> I clearly stated I was speaking a WG member and have just
>>>> as much right as anyone else to state my opinion as to whether or
>>>> not a specific piece of work is accepted by the WG. If you have
>>>> technical comments related to the draft as opposed to critiques of
>>>> character, I'd encourage you to share them.
>>>>
>>>> Thanks,
>>>> Acee
>>>>
>>>>         
>>>>> It reminds me of  folks at the ATM forum in the past !!
>>>>> Thanks,
>>>>> Abhay
>>>>>
>>>>>
>>>>>
>>>>> Acee Lindem wrote:
>>>>>           
>>>>>> Speaking as a WG Member and author of the RFC 2740 respin:
>>>>>>
>>>>>> My opinion is that the protocol extensions described in the draft
>>>>>> are completely unnecessary and the OSPFv3 (RFC 2740 and its
>>>>>> update) already can support anycast addresses as /128 prefixes.
>>>>>> One simply uses the lowest cost path. Note that the protocol
>>>>>> already support configuration of host routes as described in
>>>>>>             
> appendix
>   
>>>>>> C.7.
>>>>>>
>>>>>> Thanks,
>>>>>> Acee
>>>>>>
>>>>>> Acee Lindem wrote:
>>>>>>             
>>>>>>> Speaking as WG Chair:
>>>>>>>
>>>>>>> I'd like to initiate discussion on this draft so that we can let 
>>>>>>> the authors
>>>>>>> know whether the WG wants to presue it furthers.
>>>>>>>
>>>>>>> Thanks,
>>>>>>> Acee
>>>>>>>
>>>>>>> Yue Wang wrote:
>>>>>>>               
>>>>>>>> This document describes AOSPF (Anycast Extensions to OSPFv3),
>>>>>>>>   a routing protocol which extends OSPFv3 to support anycast,
>>>>>>>>                 
> which
>   
>>>>>>>>   we implemented and tested successfully in our IPv6 test bed.
>>>>>>>>                 
> And
>   
>>>>>>>>   the performance analysis shows the overhead of AOSPF is low.
>>>>>>>>
>>>>>>>> http://www.ietf.org/internet-drafts/draft-wang-aospf-00.txt
>>>>>>>>
>>>>>>>>
>>>>>>>>                 
>>>>>>> _______________________________________________
>>>>>>> OSPF mailing list
>>>>>>> [e-mail]
>>>>>>> https://www1.ietf.org/mailman/listinfo/ospf
>>>>>>>
>>>>>>>               
>>>>>> _______________________________________________
>>>>>> OSPF mailing list
>>>>>> [e-mail]
>>>>>> https://www1.ietf.org/mailman/listinfo/ospf
>>>>>>
>>>>>>
>>>>>>             
>>>> _______________________________________________
>>>> OSPF mailing list
>>>> [e-mail]
>>>> https://www1.ietf.org/mailman/listinfo/ospf
>>>>
>>>>
>>>>         
>>     
>
>   

_______________________________________________
OSPF mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/ospf