Re: [netlmm] I-D Action:draft-ietf-netlmm-lma-discovery-05.txt

[jouni korhonen <[e-mail]>]

Inline..


On Sep 14, 2010, at 10:37 PM, Behcet Sarikaya wrote:

> Hi Jouni,
> 
> 
> 
> ----- Original Message ----
>> From: jouni korhonen <[e-mail]>
>> To: Behcet Sarikaya <[e-mail]>
>> Cc: Vijay Devarapalli <[e-mail]>; [e-mail]
>> Sent: Tue, September 14, 2010 11:43:27 AM
>> Subject: Re: [netlmm] I-D Action:draft-ietf-netlmm-lma-discovery-05.txt
>> 
>> Hi Behcet,
>> 
>> 
>> On Sep 14, 2010, at 6:52 PM, Behcet Sarikaya  wrote:
>> 
>>> 
>>> ----- Original Message ----
>>>> From: Vijay  Devarapalli <[e-mail]>
>>>> To: Behcet  Sarikaya <[e-mail]>
>>>> Cc: Behcet  Sarikaya <[e-mail]>; jouni  korhonen 
>>>> <[e-mail]>; [e-mail]
>>>> Sent: Mon,  September 13, 2010 11:09:36 PM
>>>> Subject: Re: [netlmm] I-D  Action:draft-ietf-netlmm-lma-discovery-05.txt
>>>> 
>>>> On  9/13/10 3:49 PM, Behcet Sarikaya wrote:
>>>>> Hi  Jouni,
>>>>>   Thanks for updating the  draft.
>>>>> Reading it I noticed that in  Section 3.3on  Constructing the LMA FQDN from 
>> 
>> a
>>>>> Service   Name
>>>>> 
>>>>> 
>>>>> 
>>>>> can you  please mention a service name in  QNAME format:
>>>>> "_pmip6_ipv6.example.com"
>>>>> where example.com can come  from MN's  NAI or the operator domain.
>>>> 
>>>> No, PMIPv6  is not a service. So we shouldn't  do this.
>>> 
>>> Why not? mip6  IS a service name. We do need pmip6 service name as well, 
>> don't 
>> 
>>> you  think?
>>> 
>>> Also reading the draft's Section 3.3 on service name  (Constructing the LMA 
>> FQDN 
>> 
>>> from a Service Name to be precise), it  totally sounds like a 3GPP document.
>>> Remember this is IETF work. IETF  defined DNS and PMIPv6. 
>> 
>> I assume you did not notice the references that  were given as examples in 
>> Section 3.3? They totally seem to point in 3GPP's  general direction.
> 
> 
> Of course I did. I know that 3GPP does favor DNS discovery, in almost everything 
> they use DNS discovery. 
> 
>> 
>>> IETF can define a service name which is operator  and SDO independent. If 
>> such a 
>> 
>>> service name is defined, there is even  no need to standardize it in other 
>> SDOs 
>> 
>>> like 3GPP. Operators can use it  immediately.
>> 
>> IETF can.. but does not have to. I like the current content  of Section 7. For 
> 
>> the record, the 3GPP boat has already sailed. Not much to do  there.
>> 
> 
> I guess you mean 
> 19.4.3    Service and Protocol service names for 3GPP 
> in 23.003. Yes they defined service names for pmip like 3gpp-pgw:x-s8-pmip 
> indicating s8 interface on a P-GW.
> 
> Of course this is good and maybe you should specifically mention this in Sec. 
> 3.3.

I don't really see a reason for that. Especially when the text in 3.3 refers to APNs as an explicit example of service names rather than S-NAPRT app-service/protocol.

> 
> However the above service name is 3GPP specific. I think it is advisable to 
> define an IETF service name so that everybody can use it.

You are now mixing stuff here.. service as a identifier for network behind a gateway is different than a role/capability of a box.

> 
> These are my points.

.

> 
> 
> Regards,
> 
> Behcet
> 
> 

- Jouni


>