Re: [CCAMP] clarification about draft-takacs-ccamp-revertive-ps

[Francesco Fondelli <[e-mail]>]

On Thu, Apr 4, 2013 at 4:49 PM, BRUNGARD, DEBORAH A <[e-mail]> wrote:
> Hi Francesco,

Hi Deborah,

> That was a fast analysis:-)
>
> Suggest you first expand on the requirements that you want to support vs. solution. When you detail aspects of SNC/N, SNC/I, and SNC/S, you will find that you also need OAM configuration to support (N/I/S configuration (e.g. DEG, TTI)) and you will need to configure which defects contribute to a "failure". This will all need to be coordinated with the LSP OAM configuration.

  SNC sub-type has been introduced only recently in [4] and is a
Zafar Ali's contribution; I think he can expand/elaborate this
topic here on the list or in the next draft version.  In my emails
I was referring exclusively to hold off and wtr ([4] is about hold off
and wtr since day zero: July, 2008), sorry if this was not clear.

> Also on SNC/N, SNC/I, and SNC/S, is this for a segment or e-2-e? The draft needs more detail and alignment with GMPLS protection terminology and mechanisms.

  I think [4] addresses both e2e and segment (note: I'm talking about
hold off and wtr).

> After having a better scope of the requirements, we can discuss the solution tradeoffs. There are always multiple ways to solve, first we need to understand what is needed.

  The requirement is straightforward: in order to set up protection
switching you need hold off and wtr.  Either you use default values
for any LSP on a given network (which is what is happening today) or
you signal these values (hoff and wtr) on a per LSP basis.  The draft
tries to provide support for the latter.

> Thanks,
> Deborah

thank you
Ciao
Fra

> -----Original Message-----
> From: Francesco Fondelli [mailto:[e-mail]]
> Sent: Wednesday, April 03, 2013 12:59 PM
> To: BRUNGARD, DEBORAH A
> Cc: [e-mail]
> Subject: Re: [CCAMP] clarification about draft-takacs-ccamp-revertive-ps
>
> Hi Deborah, all,
>
>>   Having said that, I have no problem rewriting [4] using OAM
>> configuration TLV.  It's just weird to me but I can live with it.
>
>   sorry, I changed my mind, I cannot use the OAM TLV.  The more I read
> about OAM in IETF the more I think protection switching provisioning
> is completely out-of-scope.
>
>   I spent some hours looking for the OAM definition within the
> IETF context(s).  The most recent and enlightening (to me at least)
> documents I found are [A] and [B].  As far as I understand, [1] is
> perfectly aligned with them.  At the same time I cannot find any
> support of your statement:
>
>> Protection switching provisioning has always been treated as a
>> common equipment management functionality [cut]. So it is in scope
>> of OAM configuration.
>
>   Maybe I'm just missing something big (?) Can someone shed some light on
> this?
>
> thank you
> ciao
> fra
>
> [A]
> http://tools.ietf.org/html/draft-ietf-opsawg-oam-overview-08
>
> [B]
> http://tools.ietf.org/html/rfc6291
>
> On Tue, Apr 2, 2013 at 11:22 AM, Francesco Fondelli
> <[e-mail]> wrote:
>> On Mon, Apr 1, 2013 at 8:06 PM, BRUNGARD, DEBORAH A <[e-mail]> wrote:
>>> Hi Francesco,
>>
>> Hi Deborah,
>>
>>> While these may be protection switching parameters, this draft is about configuration of these parameters. Protection switching provisioning has always been treated as a common equipment management functionality - same as performance management and fault management (refer to G.7710 section 8). So it is in scope of OAM configuration. CCAMP's OAM configuration work has been focused on PM and FM but it is generally applicable (hopefully) to any equipment management configuration.
>>
>>   Puzzled.  If we follow this reasoning (i.e. common equipment management
>> functionalities => should use OAM framework) then almost any aspect of
>> networking can be applicable to OAM and so any operation should exploit
>> the OAM framework draft.
>>
>>   For example, G.7710 section 8.6.1 describes the provisioning
>> of cross-connections but this does not imply that we are going to use the
>> OAM framework to establish label binding in the next GMPLS controlled
>> technology, I guess we will continue to use LABEL_REQUEST/LABEL objects
>> (plus any other relevant info).
>>
>>> Lou's comment is that the WG has chosen the approach used in the OAM framework document for configuration. Instead of updating the protection object at this time as your draft proposes, the question is have you considered using the OAM configuration TLV? First, we need to understand why you have chosen to not use this approach. Then we can discuss pros and cons.
>>
>>   Well, at the beginning we did not take it into consideration
>> because [4] predate [1].  Later we did not take [1] into consideration
>> simply because we thought [4] was out of OAM framework scope.
>>
>>   Having said that, I have no problem rewriting [4] using OAM
>> configuration TLV.  It's just weird to me but I can live with it.
>>
>>> BR-
>>> Deborah
>>
>> thank you
>> ciao
>> fra
>>
>>>
>>> -----Original Message-----
>>> From: [e-mail] [mailto:[e-mail]] On Behalf Of Francesco Fondelli
>>> Sent: Monday, April 01, 2013 12:20 PM
>>> To: [e-mail]
>>> Subject: [CCAMP] clarification about draft-takacs-ccamp-revertive-ps
>>>
>>> quoting item 15, from www.ietf.org/proceedings/86/minutes/minutes-86-ccamp
>>>
>>> Lou Berger: I think you misunderstood my comment from the last meeting. You
>>> should look at leveraging the OAM configuration work which came after the
>>> earlier versions of your draft.
>>> Zafar Ali: this is applicable to multiple technologies.
>>> Lou Berger: yes, the OAM configuration framework is also applicable to
>>> multiple technologies. You need a strong reason not to follow the WG in
>>> this area. Please look at the OAM configuration document
>>> [draft-ietf-ccamp-oam-configuration-fwk] and either follow it or state why
>>> your work is justified in not following the existing WG solution in this
>>> area.
>>>
>>> ---
>>>
>>> Hi all,
>>>
>>>   the OAM configuration framework [1] is about OAM.  Therefore, it is used in
>>> order to signal OAM functionalities: CC/CV and PM/FM in MPLS-TP [2], CC/CV
>>> TTI/SAPI/DAPI in SONET/SDH/OTN [3]... while our draft [4] is about *protection
>>> switching*.  HOFF, WTR and SNC sub-type are protection switching parameters.
>>>
>>>   I believe HOFF, WTR and SNC sub-type are outside of the OAM configuration
>>> framework scope and should be signaled as any other protection switching
>>> params (i.e. via PROTECTION object).
>>>
>>>   I hope this answer Lou question reported above (item 15, IETF 86 ccamp
>>> minutes).  Authors of [4] would like to understand WG's view about this point
>>> and are soliciting for comments.
>>>
>>> thank you
>>> ciao
>>> FF
>>>
>>> [1]
>>> http://tools.ietf.org/html/draft-ietf-ccamp-oam-configuration-fwk-09
>>>
>>> [2]
>>> http://tools.ietf.org/html/draft-ietf-ccamp-rsvp-te-mpls-tp-oam-ext-11
>>>
>>> [3]
>>> http://tools.ietf.org/html/draft-ietf-ccamp-rsvp-te-sdh-otn-oam-ext-05
>>>
>>> [4]
>>> http://tools.ietf.org/html/draft-takacs-ccamp-revertive-ps-08
>>> _______________________________________________
>>> CCAMP mailing list
>>> [e-mail]
>>> https://www.ietf.org/mailman/listinfo/ccamp