RE: [EFM] Changes to 100BASE-X and 1000BASE-X PCS, 10G RS
With great fear I here tread....
As we prepare for this discussion a 2 weeks, let us remember that one of the
principal reasons for standardizing 100M and 1G optics at 10km is that the
parts available from the industry (when we started) could not be assured to
be mutually interoperable. In many cases, under many conditions, they were
probably interoperable. To create a standard that assures backward
interoperability with parts that were themselves
not-necessarily-interoperable... well, I think that this might be an effort
in futility.
In short, even if we choose to remove the OAM requirements, there is no
assurance that even with the an identical PCS that the parts will "play
nice."
For my part, if we were going to allow OAM to be an option, I would strongly
prefer that it be an option like 802.3ad, which a customer can readily
identify as a supported feature on a spec sheet, and not an option that is
buried in a PIC table and not readily exposed to the buyer. Yes, I realize
that 802.3ad was a project, not a clause. Yes, I understand that doing
anything like this with OAM is not possible at this stage. That would have
required a separate project.
jonathan
> -----Original Message-----
> From: owner-stds-802-3-efm@majordomo.ieee.org
> [mailto:owner-stds-802-3-efm@majordomo.ieee.org]On Behalf Of Howard
> Frazier
> Sent: Wednesday, March 03, 2004 10:38 AM
> To: stds-802-3-efm@ieee.org
> Subject: RE: [EFM] Changes to 100BASE-X and 1000BASE-X PCS, 10G RS
>
>
>
>
> Forwarded from Piers Dawe.
>
> Subject: RE: [EFM] Changes to 100BASE-X and 1000BASE-X PCS, 10G RS
> Date: Tue, 2 Mar 2004 15:07:10 -0000
> From: <piers_dawe@agilent.com>
> To: <benjamin.brown@ieee.org>, <stds-802-3-efm@ieee.org>,
> <stds-802-3@ieee.org>
> Ben,
>
> This is not JUST a project for the access network and that is not the
> "whole reason they exist". 100BASE-LX10 and 1000BASE-LX10 like PHYs
> existed before EFM, and we should be standardizing them
> right. We have
> known all along that they have general applicability. Remember, 100
> Mb/s on SMF started as a separate call for interest and was
> rolled into
> EFM for synergy.
>
> The EFM "environment" is not so different. It's the same frames, same
> rates, same wavelength, same fiber type as "legacy" 1000BASE-LX and
> 10GBASE-L. Same optional OAM proposed for all. Interoperable and
> interchangeable PMDs. So why would the PCS be different?
>
> I don't believe that the proposed mandatory PHY changes are
> "particularly tuned" even for the access market and I don't see your
> "less applicable - more applicable" trade off. By demanding
> currently
> non-standard behavior they go against Broad Market Potential,
> Compatibility and Economic Feasibility even for the access
> market. They
> make it harder to connect a "legacy Ethernet" data backbone
> network to a
> not-quite-Ethernet "EFM" access network. Do NEMs have to make boxes
> where some long wavelength GBIC ports have one PCS behavior and other
> long wavelength GBIC ports have the opposite behavior? I suppose a
> service provider can go to ATM and back to join the two!
>
>
> Let's quote from the 100BASE-FX over dual Single Mode Fibre Call For
> Interest of two years ago. Remember, as the web site
> http://www.ieee802.org/3/smfx_study/index.html says,
> The 100BASE-FX over dual Single Mode Fibre Call For Interest
> resulted in
> additional work being added to IEEE P802.3ah Ethernet in the
> First Mile
> task Force.
>
> >From EFM minutes
> <http://www.ieee802.org/3/efm/public/mar02/minutes_03_2002.pdf>
> http://www.ieee802.org/3/efm/public/mar02/minutes_03_2002.pdf
> Additional Objective: p2p 100Mb/s on SM fiber
> Bruce T. presented a motion:
> To add an objective to the family of physical layer specifications
> 100Base-X >3D 10 km over SM fiber
> ALL - for 105; Against 4; Abstained 22
> 802.3voters - for 59; Against 3; Abstained 9
> Motion passed
>
> And from 802.3 minutes
> <http://www.ieee802.org/3/minutes/mar02/minutes_0302.pdf>
> http://www.ieee802.org/3/minutes/mar02/minutes_0302.pdf :
> 802.3ah motion #1
> Add an objective to the family of physical layer specifications:
> 100BASE-X >3D 10 km over SM Fiber
> All Y:105 N: 4 A:22
> .3 Y:59 N: 3 A: 9
> Motion Passed
>
> And these quotations below (my emphasis) are from the20
> 100 Mb/s over Dual SM Fiber 100 Mb/s over Dual SM Fiber
> Proposed PAR & 5 Criteria Proposed PAR & 5 Criteria
> <http://www.ieee802.org/3/smfx_study/public/jonsson_1_0302.pdf>
> http://www.ieee802.org/3/smfx_study/public/jonsson_1_0302.pdf
>
> Scope:
> - Make amendment to Clause 26, 100BASE-FX, to include a
> 100Mbps dual SMF
> PMD
>
> Broad Market Potential Broad Market Potential
> ...
> 100BASE-X SMF is main candidate for volume applications in:
> - Residential (FTTH)
> - Commercial (SME, Shopping malls, etc.)
> - Industrial ( <http://ethernet.industrial-networking.com>
> http://ethernet.industrial-networking.com)
> o Rapid growth anticipated in emerging areas
> - fiber to the radio base stations (FTTR)
> - fiber to WLAN HotSpots (FTTW)
> - fiber links connecting office desktops (FTTD)
>
> Compatibility
>
> 100BASE-X PCS & PMA assumed, and the 802.3 MAC
> - No changes whatsoever to the MAC
> - PHY identical to current 100Mbps Std except for a new PMD
> - No change to Clause 24
> - Retain all state machines, 4B/5B coding etc. of 100BASE-X
> o Only need to extend Clause 26, 100BASE-FX PMD, to include SMF
> o Physical medium compatibility through SMF
> - Compatible with existing 1000BASE-LX
> - Provides upgrade paths to higher speeds and multiple
> wavelengths, with
> fiber plant untouched
>
> - 100Mbps optical SMF components exist
> - 'Pre-standard' links and systems already in commercial operation
>
> 100Mbps and EFM
> o EFM deals with major additions to the 802.3 Std
> o 100BASE-X dual SMF only requires minimal additions to Clause 26
> o 100BASE-X dual SMF is already happening, and will have applicability
> even outside EFM
> o However, 100BASE-X SMF will be used in the public access application
> space
> o 100BASE-X PCS is transparent to EFM OAM
> - Neither "OAM in Frames" nor "OAM on Preamble" require any changes to
> 100BASE-X PCS
>
> Piers
>
>
>