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

RE: [10GBASE-T] Proposed modification to PAR scope




From Chris Di Minico...

-----Original Message-----
From: CDimi80749@aol.com [mailto:CDimi80749@aol.com] 
Sent: Thursday, September 25, 2003 12:24 PM
To: Booth, Bradley; stds-802-3-10gbt@ieee.org
Subject: Re: [10GBASE-T] Proposed modification to PAR scope


Stuart,

In response to your comment, and mostly for the benefit of those
untrained in 
the art of cabling, 
I'll attempt to clarify the points in the discussion by providing
information 
on the ISO nomenclature
to help reach closure on the cabling language used in the objectives. 


1. ISO/IEC 11801  specifies horizontal cabling and backbone cabling
(i.e., 
topology). 

2. ISO/IEC 11801 specifies both balanced cabling (copper) and optical
fibre 
cabling (i.e., media).
------For balanced cabling,  the performance specifications are
separated 
into six classes (A to F) 
------For Class D, E and F balanced components are defined in terms of
their 
category. 
------The category of component provides the Class of balanced cabling 
performance.
+++++e.g., Category 6 components provide Class E balanced cabling
performance

3. ISO/IEC specifies transmission performance (topology) of the cabling
in 
terms of the Channel (and permanent link).


 To address each cabling attribute in a single objective we would need
to 
describe each and then 
select an ordering of the attributes to generate the objective. 
e.g., 
+++Horizontal (topology) 
+++Balanced (media) 
+++Class/Category (transmission performance specifications) 
+++Channel (transmission performance topology). 

The resultant objective would read: 
10GBASE-T Cabling: ====== Horizontal Balanced Class (D,E,F) Channel 

At this point, I do not advocate modifying the objectives. I offer this
to 
clarify 
the points of discussion concerning the applicability of the usage of 
horizontal cabling for 
data center topologies. 


Regards,


Chris


In a message dated 9/25/03 11:02:10 AM Eastern Daylight Time, 
bradley.booth@intel.com writes:

<< This is being forward on Stuart's behalf as his restrictive notice
 caused the email to be bounced.  Thanks, Brad
 
 
------------------------------------------------------------------------
 
------------------------------------------------------------------------
 -----------------------
 
 Dear all,
 
 Neither the TIA - where horizontal cabling extends from the
 cross-connect to the TO - nor ISO/IEC - where horizontal cabling
 subsystem is the equivalent - appears to be correct.
 
 Is it not the "horizontal balanced cabling channel" or the "balanced
 cabling horizontal channel"?
 
 Regards 
 
  
 
 SJReeves 
 
 Stuart J. Reeves
 
 Technical Manager 
 
 _________________________________________ 
 
 YOUR NETWORKS - OUR STRENGTH! 
 
 KRONE UK celebrating 25 Years of making the right connections 
 
 _________________________________________ 
 
 KRONE (UK) Technique Limited
 
 Runnings Road
 
 Kingsditch Trading Estate
 
 Cheltenham
 
 Gloucestershire 
 
 GL51 9NQ
 
 U.K.
 
 Email: sreeves@krone.co.uk < mailto:sreeves@krone.co.uk
 <mailto:sreeves@krone.co.uk>  <mailto:sreeves@krone.co.uk
 <mailto:sreeves@krone.co.uk> > >
 
 Tel: +44 (0)1242 264 471
 
 Fax: +44 (0)1242 264 652
 
 Mobile: +44 (0)7768 463 265 
 
    -----Original Message-----
    From: owner-stds-802-3-10gbt@majordomo.ieee.org
 [mailto:owner-stds-802-3-10gbt@majordomo.ieee.org] On Behalf Of Cobb,
 Terry R (Terry)
    Sent: Thursday, September 25, 2003 9:06 AM
    To: stds-802-3-10gbt@ieee.org
    Subject: RE: [10GBASE-T] Proposed modification to PAR scope
    
    
    The term ISO uses is "Balance Cabling" (Clause 6 of 11801, 2002)
 and TIA uses "Horizontal Cabling", which is used in the Data Center
 document to identify the cabling. I have been reminded often by others
 that we should use ISO designations.
     
    Terry
 
        -----Original Message-----
        From: pat_thaler@agilent.com
 [mailto:pat_thaler@agilent.com]
        Sent: Wednesday, September 24, 2003 6:26 PM
        To: gzimmerman@solarflare.com; dan.dove@hp.com;
 btolley@cisco.com; bradley.booth@intel.com; stds-802-3-10gbt@ieee.org
        Subject: RE: [10GBASE-T] Proposed modification to PAR
 scope
        
        
        I agree that "on horizontal structured cabling" probably
 needs a tweak. It could possibly be interpreted as requiring 100 m
 operation for all supported media. As Terry points out, it doesn't
 really reflect the data center as a focus of broad market potential.
         
        Regards,
        Pat
 
            -----Original Message-----
            From: George Zimmerman
 [mailto:gzimmerman@solarflare.com]
            Sent: Wednesday, September 24, 2003 3:42 PM
            To: THALER,PAT (A-Roseville,ex1);
 dan.dove@hp.com; btolley@cisco.com; bradley.booth@intel.com;
 stds-802-3-10gbt@ieee.org
            Subject: RE: [10GBASE-T] Proposed modification
 to PAR scope
            
            
 
            Pat - thanks for the clarifications - I'd missed
 the fact that speed wasn't mentioned (must be still suffering from
 Italian jet-lag).  Scopes should be broad but clear.  The CX4 scope is
 probably a good model.  "based on" is different than saying it must
 implement it.  (under this scope a 10GBASE-CX4 could incorporate other
 line codes, etc.)  Similarly "working over the wiring types used in
 structured cabling" is a bit different than the text as written, which
 enters into a more specific description of "structured cabling" (we had
 a little discussion in Italy where some had a very narrow understanding
 of what that means). 
 
             
 
            I'll have to think a little about an
 alternative, but I think we're on the same principle: speed & wiring
 types define 10GBASE-T, but the detailed description is for the
 objectives.
 
             
 
             
 
            George Zimmerman
 
            gzimmerman@solarflare.com
 
            tel: (949) 581-6830 ext. 2500
 
            cell: (310) 920-3860
 
            -----Original Message-----
            From: pat_thaler@agilent.com
 [mailto:pat_thaler@agilent.com] 
            Sent: Wednesday, September 24, 2003 3:28 PM
            To: George Zimmerman; dan.dove@hp.com;
 btolley@cisco.com; bradley.booth@intel.com; stds-802-3-10gbt@ieee.org
            Subject: RE: [10GBASE-T] Proposed modification
 to PAR scope
 
             
 
            George,
 
             
 
            I think Dan's suggestion went too far - very
 specific media descriptions (like references to 11801) have always been
 for the objectives rather then the PAR. On the other hand, the existing
 scope is much more broad than previous projects:
 
             
 
            Current scope:
 
            The scope of this project is to specify
 additions to and appropriate modifications of IEEE Std 802.3 (including
 all approved amendments and corrigenda) to add a copper Physical Layer
 (PHY) specification.
 
             
 
            Howard's suggested scope:  
 
            Specify a Physical Layer (PHY) for operation at
 10 Gb/s 
               on horizontal structured copper cabling,
 using the existing 
               Media Access Controller, and with extensions
 to the appropriate 
               physical layer management parameters, of IEEE
 Std 802.3
 
             
 
            CX4 scope:
 
            The scope of this project is to specify
 additions to and appropriate modifications of IEEE Std 802.3 as amended
 by IEEE Std   802.3ae-2002 (and any other approved amendment or
 corrigendum) to add a copper Physical Medium Dependent (PMD) option for
 10   Gb/s operation, building upon the existing 10GBASE-X Physical
 Coding Sublayer (PCS) and 10 Gigabit Attachment Unit Interface   (XAUI)
 specifications. 
 
             
 
            The CX4 scope text is much more similar to
 Howard's suggested scope. Both have a statement about the speed. I
can't
 recall any scope statement we have done for a PHY project that omitted
 mention of speed. The CX4 scope doesn't say anything about the type of
 copper, but it specifies that the PHY will be based on the X PCS and
the
 XAUI specs which limits it pretty clearly. For 10GBASE-T, the intent to
 work over the wiring types used in structured cabling and the 10 Gbit/s
 speed are the defining factors.
 
             
 
            Look at it this way. IEEE Std 802.3 already has
 many copper Physical Layer specifications. Therefore the job listed in
 the current scope statement has already been done. If the PAR is
 approved with the current scope, the scope will be published by the
 IEEE. How would a reader seeing that scope know what the project was
 about and whether they were interested? Howard's scope is a more clear
 statement of what we want to do.
 
             
 
            If something in Howard's scope is too confining,
 then please propose an alternative that is reasonably descriptive of
the
 particular nature of this project - not something that could describe 5
 or more other projects we have already done.
 
             
 
            Regards,
 
            Pat