RE: [EFM] OAM compromise
I would assert that as far as the business subscribers of our xLEC customers
are concerned 1GE will become the equivalent of today's T1/E1 access
circuits, operating p2p, probably over the same fiber that is used today for
T3 or E1 delivery. See below for the supporting arguments please.
We also have the unbundling / wholesale issue to cover, again, more of an
issue in Europe than it is in the US, at this time.
I support having OAM preamble for fast signalling and constant monitoring,
and OAM frames for sending data. Two different problems, therefore it is
acceptable to have two solutions.
This also maps to the copper track where the OAM transport for fast
signalling and constant monitoring functions are already in the PHYs. Copper
does not need to use the pre-amble, and benefits from only sending the MAC
frame with no pre-amble or IPG.
So my compromise would be:
Phy layer OAM information transported over pre-amble in 1GE and EPON, and
whatever management channel there is in whatever PHY the copper track use.
Layer two OAM information transported over OAM frames transport.
All the above to be mandatory. However, that means mandatory on EFM 1GE and
EFM EPON. It does not mean mandatory on legacy PHYs and MACs. They can use
OAM in frames.
Done deal :-).
Here is the background stuff:
In Europe there is a lot of E1 delivery over fiber, and where there is no
fiber there are ducts with draw wires. This changes the cash flow costs in
the business model (even if there is 'funny money' in the business case).