Re: [Detnet-dp-dt] DP DT solution draft slides for IETF99

[Loa Andersson <[e-mail]>]

Folks,

I think this discussion to a large extent is moot.

 From one perspective all labels are T-Labels, sometimes we want to
discuss what these tunnels are there for.

S-Labels indicated that they are carrying some type of service, not
being particular about what type of service, it is just something we
treat as a service.

During the discussion of the of detnet replication and elimination, we
thought it would be good to create an overlay network (L-Labels) we
found that we did not need to name the L-Labels, but they are still
logically there.

On 2017-07-14 14:35, Jiangyuanlong wrote:
> Thanks much, Lou.
> It seems that if and only if multiple detnet services are aggregated into a tunnel (T-Label), the S-Label is needed to distinguish them; otherwise, the S-Label is not needed.
> Correct?

Maybe - are you implying that that we could prohibit multiple detnet
services being aggregated into a tunnel? I think not! S we need the 
concept of an S-Label.

I think that it is good practice to have S-Labels if we have multiple
instances of the same service multiplexed on the same T-Label, right?

/Loa
>
> Cheers,
> Yuanlong
>
> -----Original Message-----
> From: Detnet-dp-dt [mailto:[e-mail]] On Behalf Of Lou Berger
> Sent: Friday, July 14, 2017 6:22 PM
> To: Jiangyuanlong; Jouni; [e-mail]; 'Korhonen, Jouni'; 'Balázs Varga A'; [e-mail]
> Subject: Re: [Detnet-dp-dt] DP DT solution draft slides for IETF99
>
> T-label imo.
>
> Lou
>
>
> On July 14, 2017 4:26:06 AM Jiangyuanlong <[e-mail]> wrote:
>
>> Sorry, I was confused by L-label in the last version and S-Label. But
>> we still need to harmonize the T-Lable with the S-Label.
>> For example, if we set up a low-latency or contention-free LSP for a
>> detnet flow (between DA-T-PEs or DA-S-PEs), most probably we need some
>> traffic engineered LSPs (i.e., L-LSP as defined in RFC 3270).
>> Can we regard L-LSP labels on the path to be a T-Label or an S-Label?
>>
>> Best regards,
>> Yuanlong
>>
>> -----Original Message-----
>> From: Detnet-dp-dt [mailto:[e-mail]] On Behalf Of
>> Jiangyuanlong
>> Sent: Friday, July 14, 2017 3:56 PM
>> To: Jouni; [e-mail]; 'Korhonen, Jouni'; 'Balázs Varga A';
>> [e-mail]
>> Subject: Re: [Detnet-dp-dt] DP DT solution draft slides for IETF99
>>
>> Hi folks,
>>
>> I am not sure we need to introduce S-Label in the first place.
>> As I remember, we had some consensus that PW label has carried enough
>> information in the f2f discussion happened during the last IETF meeting.
>> And S-label is regarded redundant for PW. Did I miss something?
>>
>> Thanks,
>> Yuanlong
>>
>> -----Original Message-----
>> From: Detnet-dp-dt [mailto:[e-mail]] On Behalf Of
>> Jouni
>> Sent: Friday, July 14, 2017 1:17 PM
>> To: [e-mail]; 'Korhonen, Jouni'; 'Balázs Varga A';
>> [e-mail]
>> Subject: Re: [Detnet-dp-dt] DP DT solution draft slides for IETF99
>>
>> Thanks.
>>
>>> -----Original Message-----
>>> From: Detnet-dp-dt [mailto:[e-mail]] On Behalf
>>> Of Carlos Jesús Bernardos Cano
>>> Sent: Thursday, July 13, 2017 20:39 PM
>>> To: Korhonen, Jouni <[e-mail]>; Balázs Varga A
>>> <[e-mail]>; [e-mail]
>>> Subject: Re: [Detnet-dp-dt] DP DT solution draft slides for IETF99
>>>
>>> Hi Jouni,
>>>
>>> Thanks for preparing this. Some small comments below:
>>>
>>> - Slide 6: I'd remove "native" in "PW-based native DetNet" to be
>>> consistent with the terms used in the draft (alternatively, I'd use
>>> "IPv6- based native DetNet" in slide 7for consistency with "PW-based
>>> native DetNet in slide 6).
>>
>> Oops. Good catch.
>>
>>>
>>> - Slides 11 and 12: use the same order for "Flow-ID" and "SeqNum" on
>>> the slides (right hand side)
>>
>> Ok.
>>
>>
>>> - Slide 11: though I have no concrete proposal, I think the S-label
>>> could be better introduced (maybe with a figure, also introducing the
>>> (DA-)T-PE and (DA-)S-PE node terminology).
>>
>> Ok. I'll come up with something.
>>
>>
>>>
>>> - Slide 14: "already be seen" --> "already been seen"
>>
>> Ok.
>>
>> - Jouni
>>
>>
>>
>>>
>>> Thanks,
>>>
>>> Carlos
>>>
>>> On Thu, 2017-07-13 at 13:57 +0000, Korhonen, Jouni wrote:
>>>> An update.. I am still doing the QoS etc part of the deck.
>>>>
>>>>
>>>> - Jouni
>>>>
>>>>
>>>>
>>>>> -----Original Message-----
>>>>> From: Balázs Varga A [mailto:[e-mail]]
>>>>> Sent: Thursday, July 13, 2017 12:54 PM
>>>>> To: Korhonen, Jouni <[e-mail]>;
>>>>> detnet-dp-dt@ie tf.org
>>>>> Subject: RE: DP DT solution draft slides for IETF99
>>>>>
>>>>> Hi Jouni,
>>>>>
>>>>> Thanks for preparing this. Just some fast reactions:
>>>>> - slide5-6-7: we may receive a comment that it looks like only
>>>>> end- hosts having same type (TSN, MPLS, IPv6) can be interconnected.
>>>>> I would propose to add a note, that other combinations as
>>>>> depicted requires further considerations.
>>>>>
>>>>> - slide8: we have used the MS-PW analogy during our discussions.
>>>>> However
>>>>> it is valid only if PREF is used.
>>>>> I would propose to refer on the first bullet only to "PseudoWires"
>>>>> and
>>>>> "IPv6" as the two data plane solution.
>>>>> A further note could highlight the MS-PW analogy for PREF scenarios.
>>>>>
>>>>> - slide10-11: I would pair the DetNet flow specific information
>>>>> fields to be transported with the data plane encapsulation fields.
>>>>> 	DetNet flow	Encapsulation fields
>>>>> 	Flow ID:	PW label
>>>>> 	Seq. number: 	CW
>>>>>
>>>>> - slide14: regarding multicast DetNet flows I would formulate
>>>>> somewhat different. In my view we have considered p2p data plane
>>>>> solutions.
>>>>> The defined data plane works for DetNet flows having multicast
>>>>> dst- address assuming that the DetNet domain provides p2p connectivity.
>>>>> We may also receive comments that many DetNet flows are multicast
>>>>> (e.g., TSN flows using IEEE-FRER, etc.)
>>>>>
>>>>> Cheers
>>>>> Bala'zs
>>>>>
>>>>> -----Original Message-----
>>>>> From: Detnet-dp-dt [mailto:[e-mail]] On
>>>>> Behalf Of Korhonen, Jouni
>>>>> Sent: 2017. július 12. 23:29
>>>>> To: [e-mail]
>>>>> Subject: [Detnet-dp-dt] DP DT solution draft slides for IETF99
>>>>>
>>>>> Hi,
>>>>>
>>>>> Sorry about this taking so long. Please, have a look and flame on..
>>>>> There's still time to work on the actual content. However, keep
>>>>> in mind that this is mainly an update from last time.
>>>>>
>>>>> - Jouni
>>>>
>>>> _______________________________________________
>>>> Detnet-dp-dt mailing list
>>>> [e-mail]
>>>> https://www.ietf.org/mailman/listinfo/detnet-dp-dt
>>>
>>> _______________________________________________
>>> Detnet-dp-dt mailing list
>>> [e-mail]
>>> https://www.ietf.org/mailman/listinfo/detnet-dp-dt
>>
>> _______________________________________________
>> Detnet-dp-dt mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/detnet-dp-dt
>> _______________________________________________
>> Detnet-dp-dt mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/detnet-dp-dt
>> _______________________________________________
>> Detnet-dp-dt mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/detnet-dp-dt
>
>
> _______________________________________________
> Detnet-dp-dt mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/detnet-dp-dt
> _______________________________________________
> Detnet-dp-dt mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/detnet-dp-dt
>

-- 


Loa Andersson                        email: [e-mail]
Senior MPLS Expert                          [e-mail]
Huawei Technologies (consultant)     phone: [phone]