Re: [sfc] New Version Notification for draft-merged-sfc-architecture-02.txt

[Naiming Shen <[e-mail]>]

Carlos,

On Aug 26, 2014, at 3:47 PM, Carlos Pignataro (cpignata) <[e-mail]> wrote:

> Naiming,
> 
> You are right -- this is covered in Section 4.3, under "Terminating SFPs", we can massage the definition some more.

That would be good. Although looking at the Section 4.3 the "Terminating SFPs" paragraph, it's more about the
last SFF will strip the NSH transport/header then forwarding; I'm thinking more in the area with the last SFF can be
optionally inserted just for preventing the service looping in the network and there is no SF associated with it.

thanks.
- Naiming

> 
> THanks,
> 
> Carlos.
> 
> On Aug 25, 2014, at 4:47 PM, Naiming Shen <[e-mail]> wrote:
> 
>> 
>> On Aug 25, 2014, at 1:36 PM, Naiming Shen <[e-mail]> wrote:
>> 
>>> 
>>> Carlos,
>>> 
>>> One of the function of SFF could be to "de-encapsulate" the SFC, it can be the case of the last
>>> hop on the SFC list, there may not be any SF associated with the SFF, but only to remove the service
>>> transport and NSH header and to forward the packet normally.
>>> 
>>> Otherwise, after the last SF/SFF service function, service header and transport is removed,
>>> the packet could route through the original "classifier" device again and it has no information
>>> that packet has already gone through the SFC defined. This can cause looping. This of course
>>> depends on where the location of the last SFF in the topology. One way to solve this can
>>> be to define the SFC last item being the SFF which is the next-hop of the classifier device in
>>> normal routing/forwarding to make sure the classifier device will to see this packet in original
>> 
>> typo, "… will not see this …"
>> 
>>> format(without NSH and transport) twice.
>>> 
>>> thanks.
>>> - Naiming
>>> 
>>> On Aug 25, 2014, at 12:14 PM, "Carlos Pignataro (cpignata)" <[e-mail]> wrote:
>>> 
>>>> Reinaldo,
>>>> 
>>>> Thanks for the comment, good set of points. It does seem that the definition itself might be unnecessarily overly restrictive.
>>>> 
>>>> We could say "zero or more" or we could say "typically one or more", but I think it is better to spell out the function. Here's one more comprehensive proposal:
>>>> 
>>>> Old:
>>>>    Service Function Forwarder (SFF):  A service function forwarder is
>>>>         responsible for delivering traffic received from the network to
>>>>         one or more connected service functions according to information
>>>>         carried in the SFC encapsulation.
>>>> 
>>>> New:
>>>>    Service Function Forwarder (SFF):  A service function forwarder is
>>>>         responsible for delivering traffic received from the network to
>>>>         one or more connected service functions according to information
>>>>         carried in the SFC encapsulation, as well as for delivering traffic to
>>>>         a classifier or mapping out traffic to another SFF (in the same or
>>>>         different type of overlay).
>>>> 
>>>> WG, Reinaldo,
>>>> 
>>>> Thoughts?
>>>> 
>>>> Thanks,
>>>> 
>>>> Carlos.
>>>> 
>>>> On Aug 25, 2014, at 1:06 PM, Reinaldo Penno <[e-mail]> wrote:
>>>> 
>>>>> A couple of points about SFF definition. You mention "one or more connected service functions" 
>>>>> 
>>>>> But in our implementation we have two types of SFFs that do not have SFs:
>>>>> 
>>>>> - A SFF that maps from one overlay to another, say, VXLAN to GRE
>>>>> - A SFF that only has a classifier (no SFs in itself)
>>>>> 
>>>>> Where would they fit or how to to make sure the architecture can predict their usage?
>>>>> 
>>>>> thanks,
>>>>> 
>>>>> On 8/23/14 1:47 PM, Carlos Pignataro (cpignata) wrote:
>>>>>> SFC,
>>>>>> 
>>>>>> Please find below the email notice of a new revision of draft-merged-sfc-architecture.
>>>>>> 
>>>>>> Full set of diffs from -00 (IETF90) to -02 (now) can be seen here: http://www.ietf.org/rfcdiff?url2=draft-merged-sfc-architecture-02&url1=draft-merged-sfc-architecture-00
>>>>>> 
>>>>>> We still expect further changes to the document; but we also believe that this revision captures the key points and addresses the key open items, as planned in Toronto.
>>>>>> 
>>>>>> The key objective being to create a single document basis for the SFC architecture.
>>>>>> 
>>>>>> SFC Chairs,
>>>>>> 
>>>>>> We believe that this revision fulfills the next steps agreed in Toronto (http://tools.ietf.org/agenda/90/slides/slides-90-sfc-3.pdf). 
>>>>>> 
>>>>>> This revision, while we expect changes, is now close enough that we think it makes sense for the WG to take it as the basis for the WG document to address the deliverable.
>>>>>> 
>>>>>> draft-merged-sfc-architecture-02 addressed the key points -- and we believe is ready to start a poll for adoption. Can you please initiate that WG adoption call for draft-merged-sfc-architecture-02?
>>>>>> 
>>>>>> Thanks,
>>>>>> 
>>>>>> Carlos & Joel.
>>>>>> 
>>>>>> 
>>>>>> Begin forwarded message:
>>>>>> 
>>>>>>> From: <[e-mail]>
>>>>>>> Subject: New Version Notification for draft-merged-sfc-architecture-02.txt
>>>>>>> Date: August 22, 2014 at 12:59:13 PM EDT
>>>>>>> To: Joel Halpern <[e-mail]>, Carlos Pignataro <[e-mail]>, "Joel M. Halpern" <[e-mail]>, Carlos Pignataro <[e-mail]>
>>>>>>> 
>>>>>>> 
>>>>>>> A new version of I-D, draft-merged-sfc-architecture-02.txt
>>>>>>> has been successfully submitted by Carlos Pignataro and posted to the
>>>>>>> IETF repository.
>>>>>>> 
>>>>>>> Name: draft-merged-sfc-architecture
>>>>>>> Revision: 02
>>>>>>> Title: Service Function Chaining (SFC) Architecture
>>>>>>> Document date: 2014-08-22
>>>>>>> Group: Individual Submission
>>>>>>> Pages: 26
>>>>>>> URL:            http://www.ietf.org/internet-drafts/draft-merged-sfc-architecture-02.txt
>>>>>>> Status:         https://datatracker.ietf.org/doc/draft-merged-sfc-architecture/
>>>>>>> Htmlized:       http://tools.ietf.org/html/draft-merged-sfc-architecture-02
>>>>>>> Diff:           http://www.ietf.org/rfcdiff?url2=draft-merged-sfc-architecture-02
>>>>>>> 
>>>>>>> Abstract:
>>>>>>>   This document describes an architecture for the specification,
>>>>>>>   creation, and ongoing maintenance of Service Function Chains (SFC) in
>>>>>>>   a network.  It includes architectural concepts, principles, and
>>>>>>>   components used in the construction of composite services through
>>>>>>>   deployment of SFCs.  This document does not propose solutions,
>>>>>>>   protocols, or extensions to existing protocols.
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> Please note that it may take a couple of minutes from the time of submission
>>>>>>> until the htmlized version and diff are available at tools.ietf.org.
>>>>>>> 
>>>>>>> The IETF Secretariat
>>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> _______________________________________________
>>>>>> sfc mailing list
>>>>>> [e-mail]
>>>>>> https://www.ietf.org/mailman/listinfo/sfc
>>>>> 
>>>>> _______________________________________________
>>>>> sfc mailing list
>>>>> [e-mail]
>>>>> https://www.ietf.org/mailman/listinfo/sfc
>>>> 
>>>> _______________________________________________
>>>> sfc mailing list
>>>> [e-mail]
>>>> https://www.ietf.org/mailman/listinfo/sfc
>>> 
>> 
>