RE: [EFM-OAM] Loopback question
Norman,
You raise a good point. But it seems that 802.3ah states that the OAM is defined only for p2p and emulated p2p case and AFAIK p2p doesn't need MAC learning.
Besides 802.3ah OAM has close ties to Ethernet PHY, and possibly not applicable to other PHYs (such as symbol error counts)
What you need is a WAN OAM that is being discussed in MEF and ITU.
-Shahram.
>-----Original Message-----
>From: Norman Finn [mailto:nfinn@cisco.com]
>Sent: Tuesday, June 03, 2003 11:36 AM
>To: stds-802-3-efm-oam@ieee.org
>Subject: [EFM-OAM] Loopback question
>
>
>
>General comment on the current OAM draft. Forgive me if I'm
>bringing up
>an old argument; I wasn't present, of course, when this was settled.
>
>The loopback method chosen, to do a verbatim loopback and not swap the
>MAC addresses, certainly has the advantage of simplicity for
>the passive
>station. On the other hand, unlike all of the other protocols to date
>(except for 802.3X Pause), it demands that you have a direct, physical
>link between the MACs and not a virtual one.
>
>If you look at all the virtual technologies available for
>Ethernet, past
>and present: ATM LAN Emulation, Ether-over-MPLS,
>Ether-over-L2TPv3, IETF
>VPLS, etc., most all of them involve some degree of learning
>MAC addresses.
>Echoing verbatim will leave any such technology with the wrong idea of
>what MAC address is where. It will seriously exercise any virtual
>medium's MAC address learning capability while the loopback is
>taking place.
>That's what I mean by "OAM demands physical media".
>
>Just making sure that this factor has been discussed. The inability to
>work over virtual media could limit the long-term viability of OAM.
>
>-- Norm
>