Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [EFM] [EFM-OAM] OAM Transport Proposal

Geoff, Roy,

Please allow me to detail this specific portion of the current EFM OAM
Baseline proposal which includes a means to quickly and efficiently,
impact to customer bandwidth, report fault and alarm conditions. When
such conditions occur, and frames are flowing over the link, the report
is transported via an unused portion of a frame's preamble. When such
conditions occur, and frames are not flowing over the link, the report
is transported via an 8-byte report whose format is the same as if it
had been included in the preamble of a frame. In either case, the report
is terminated at the RS and does not get to the MAC.

The proposed optional report is applicable to all optical EFM PHYs and
some copper PHYs, providing a fairly universal mechanism for report
transport. Other Ethernet variants provide similar reports in a
variant-dependent manner. For example, IEEE P802.3ae (10GbE) supports
LF/RF reports via a 4-byte ordered-set, also in the IPG. 

Best Regards,


Roy Bynum wrote:
> Geoff,
> These are the same issues that I brought up in the baseline break out
> group.  I was over ruled by the predominately OAMiP participation in that
> group.  That is the reason that I withdrew my support of the baseline
> proposal being presented at the EFMA meeting.
> Thank you,
> Roy Bynum
> At 12:47 PM 4/20/2002 -0700, Geoff Thompson wrote:
> >Roy-
> >
> >I have not examined the proposal.
> >It sounds like the proposal to send a preamble without a frame would either:
> >         1) Violate the minimum frame size of Ethernet
> >         or
> >         2) Violate the frame format requirements of Ethernet
> >My questions would then be:
> >         Why do the proposers think that this "would be Ethernet?"
> >         How do they expect to get it approved in 802.3
> >         How do they expect these entities to show up on Ethernet test
> > equipment?
> >         Have they checked it against the 5 Criteria lately?
> >
> >Geoff
> >
> >
> >At 08:38 AM 4/20/02 -0500, Roy Bynum wrote:
> >
> >>Matt,
> >>
> >>I had to withdraw my support of this baseline proposal.  The people that
> >>were in your breakout group originally were replaced by individuals that
> >>had less of a neutral agenda.  While this might be a natural evaluation,
> >>I found the group diverting from its original direction.
> >>
> >>What forced me to withdraw was the re-insertion in the baseline proposal
> >>of the new frame type shown in slide 10 as a stand alone preamble with no
> >>Ethernet frame behind it.  This has not has very much discussion in the
> >>EFM TF group as a whole, or the dot 3 voters.  This functionality was not
> >>part of the original functionality that the breakout group has originally
> >>agreed on.
> >>
> >>Several people are saying that preamble will not work properly without
> >>that non-Ethernet-frame preamble type frame.  The alternative concept is
> >>that the OAM frame will always be available to be sent, so there is never
> >>a need to send a preamble without a frame.
> >>
> >>Also, the PHY ID was re-introduced in the P2P baseline OAM as well as the
> >>P2MP, as shown in slide 9.  The breakout group had originally agreed that
> >>P2P would not support multiple network elements within the P2P link, and
> >>so the PHY ID was specific to P2MP and should not be part of the basic
> >>OAM baseline.
> >>
> >>I enjoyed working with the group that you originally invited to
> >>participate.  Thank you for the oportunity.
> >>
> >>Thank you,
> >>Roy Bynum
> >>
> >>
> >>I have the perception (I could be wrong) that more representation from
> >>the "preamble" camp participated in developing the baseline.  For example
> >>HS was there while DG was not.
> >>
> >>At 02:58 PM 4/19/2002 -0700, Booth, Bradley wrote:
> >>
> >>>Kevin,
> >>>
> >>>I have a few questions:
> >>>*       OAM in VOC/eoc is not explained in the document.  Is there a
> >>>proposal that should be referenced?
> >>>*       Do these OAM protocols assume no repeaters?  Is the OAM scheme
> >>>designed to work in half-duplex?
> >>>*       Is there a specific OAM scheme that should be used for end-to-end
> >>>(versus link-by-link) OAM messaging?  Carrier class equipment has section,
> >>>line and path, do we have something similar?
> >>>*       IEEE Std. 802.3z currently permits GbE fiber links to generate
> >>>either 7 or 8 bytes of preamble.  How does the OAM in preamble compensate
> >>>for this?
> >>>
> >>>Thanks,
> >>>Brad
> >>>
> >>>Brad Booth
> >>>Intel Platform Networking Group - Austin
> >>>bradley.booth@xxxxxxxxx <mailto:bradley.booth@xxxxxxxxx>
> >>>
> >>>                 -----Original Message-----
> >>>                 From:   Kevin Daines
> >>>[mailto:Kevin.Daines@xxxxxxxxxxxxxxxxxxxx]
> >>>                 Sent:   Friday, April 19, 2002 4:01 PM
> >>>                 To:
> >>>                 Subject:        [EFM] [EFM-OAM] OAM Transport Proposal
> >>>
> >>>                 << File: OAMtransport_041902.pdf >> All,
> >>>
> >>>                 A number of individuals have worked since the St. Louis
> >>>Meeting in March on a compromise OAM Transport proposal. We are posting the
> >>>proposal for review/comment from the larger 802.3ah Task Force.
> >>>
> >>>
> >>>
> >>>                 Kevin Daines
> >>>