Re: [Time] Issue of OAM information Information gathering Gathering from Service Function

[Gregory Mirsky <[e-mail]>]

Hi Qin,
please find my notes in-line and tagged by GIM>>.

                Regards,
                                Greg

From: Qin Wu [mailto:[e-mail]]
Sent: Monday, June 30, 2014 1:04 AM
To: [e-mail]; [e-mail]
Cc: Gregory Mirsky
Subject: RE: Issue of OAM information Information gathering Gathering from Service Function

Hi, Greg:
Thanks for comments on SFC related cases in the problem statement draft.

http://tools.ietf.org/html/draft-ww-opsawg-multi-layer-oam-01
See my reply inline below.

Greg>Service may be in form of NF instantiated in a cloud/data center. I think that in such case there would not be physical node to point to.

[Qin]:Yes, service node can be either physical node or virtual node. Do you have any proposed change here?
GIM>> I’d propose to use “Service Function” in place of “service node” as that seems to be accepted term for an element of an Service Function Chain.


Greg> For the purpose of SFC OAM location of SF is irrelevant. Location of SF is OAM information of the lower layer than SFC OAM.
[Qin]:Correct, for SFC OAM, SF location doesn’t matter, what matter is where OAM information is located and where OAM function is performed.
Any proposed change here?

GIM>> Service Functions may be applied at different network layers, not only above Layer 3. (Interesting, if SFC may include SFs at different layers.)  I’d propose to edit the first sentence of the Section 3.6.1 to
The service packets are steered through a set of service functions distributed in the network that may be applied at different layers of the network protocol stack.