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

SAN Parallel Optics applications




Paul,

To take a slight tangent off Paul's note on SAN applications and
distance objectives, the NCITS T11 standards committee, better known as
Fibre Channel, has seen it prudent to include short haul PMDs in the set
of baseline proposals for the 10 Gigabit Fibre Channel project. One of
these PMDs is the 4 channel PMD discussed over this reflector and
presented to the Task Force. 

In general, SAN applications have a significantly higher weighting
towards short-haul than long haul links than do all other networks
including the LAN. In addition, a single fiber solution might be
desirable for SANs but cost is king. I am arguing neither for a new
P802.3ae objective nor that parallel optics satisfies an existing one. I
am merely pointing out that any SAN application argument centered on 300
m and ignoring the short haul is flawed. 

Best Regards,
Rich
   
--

Paul Bottorff wrote:
> 
> Bruce:
> 
> I'm not certain the entire group sees the 300 m objective in the same way.
> Since the 300 m is not installed fiber it can be argued that it is not an
> objective for a high percent of installed fiber runs. The committee decided
> to address only 100 m installed fiber. SAN applications might require runs
> longer than 100 m and therefore may be well served by 300 m.
> 
> I don't  want to argue in favor of parallel ribbon for SAN applications.
> Though parallel ribbon will be used for rack jumpers a single fiber
> solution is a much more desirable long term answer for SANs. I believe we
> should not standardize parallel ribbon, especially at 300 m, however a 300
> m serial MMF solution for SANs does seem appealing.
> 
> Cheers,
> 
> Paul
> 
> At 04:25 PM 7/27/2000 -0700, Bruce Tolley wrote:
> 
> >Paul and Brad:
> >
> >I do not think the 300 meter objective was written with computer rooms in
> >mind.  It was written with building and campus backbones in mind.
> >
> >This was the context of the discussion within the distance ad hoc leading
> >up to the York meeting.
> >
> >My recollection from the York meeting is that we started discussing an
> >objective for 300 meters on installed fiber and ended up with consensus
> >on  objective that omitted the word "installed. "
> >
> >At the time, it seemed to me to be a compromise between those who wanted
> >to support the installed base with 1300 WWDM and those who wanted to
> >support 850 serial PMDs on new MM fiber.  Both sets of voters were
> >thinking in terms of building and campus backbones. The distance value of
> >300 meters itself has always been justified on the basis that it covers
> >some substantial percentage of the installed base of building and campus
> >backbones.
> >
> >Yours
> >
> >Bruce
> >
> >At 02:35 PM 7/27/00 -0700, Booth, Bradley wrote:
> >
> >>Paul,
> >>
> >>Thanks for pointing that out.  I stand corrected.  The parallel optics and
> >>parallel fiber could be applied to the 300m over MMF objective.  Although it
> >>does meet that objective, the previous emails were targeting parallel
> >>optics/fiber for the 100m application, and I should have probably qualified
> >>my statement with that.
> >>
> >>If it is only going to meet the 300m over MMF objective (and from Pat
> >>Gilliland's presentation, only on new high bandwidth MMF), then I have the
> >>same problem with this solution as I do with the 850nm Serial PMD solution.
> >>They may both be the lowest cost today, but they don't satisfy the 100m over
> >>installed MMF objective.  Considering we have two PMDs that are under
> >>consideration to meet both the 100m over installed MMF and 300m over MMF
> >>objectives, I believe it would be in the Task Force's best interest to focus
> >>on those solutions.  That is just my humble opinion.
> >>
> >>Cheers,
> >>Brad
> >>
> >>Brad Booth
> >>Intel LAN Access Division, Austin Design Center
> >>bradley.booth@xxxxxxxxx <mailto:bradley.booth@xxxxxxxxx>
> >>(512) 407-2135 office
> >>(512) 589-4438 cellular
> >>
> >>
> >>                 -----Original Message-----
> >>                 From:   Paul Bottorff [mailto:pbottorf@xxxxxxxxxxxxxxxxxx]
> >>                 Sent:   Thursday, July 27, 2000 3:34 PM
> >>                 To:     Booth, Bradley; stds-802-3-hssg@xxxxxxxx
> >>                 Subject:        RE: Equalization and benefits of Parallel
> >>Optics.
> >>
> >>                 Brad:
> >>
> >>                 I also understand our objectives in the same way. We don't
> >>have an
> >>                 objective for 100 m computer room connections. It seems to
> >>me the 300 m
> >>                 objective was written for computer rooms. The 300 m over MMF
> >>could be
> >>                 applied to any fiber solution.
> >>
> >>                 Cheers,
> >>
> >>                 Paul
> >>
> >>                 At 12:55 PM 7/27/2000 -0700, Booth, Bradley wrote:
> >>
> >>                 >Ali,
> >>                 >
> >>                 > From my understanding of the objectives, the task force
> >>doesn't have a
> >>                 >distance objective of "100m data center applications."  We
> >>do have an
> >>                 >objective for 100m over installed MMF fiber.  That 100m
> >>distance objective
> >>                 >was chosen because it reflects what is used in the data
> >>center applications.
> >>                 >If the task force satisfies the objective (which is a
> >>requirement for the
> >>                 >task force to do), then we provide a solution for the
> >>application.  The
> >>                 >reverse is not true.  If task force satisfies the
> >>application, then we don't
> >>                 >meet our objectives.
> >>                 >
> >>                 >Given that the task force has to satisfy objectives first
> >>and foremost, I
> >>                 >believe that it is key that the task force focus on those
> >>proposals that in
> >>                 >some manner satisfy an objective.  As I see it, parallel
> >>optics and parallel
> >>                 >fiber do not satisfy any of our objectives; therefore, the
> >>task force needs
> >>                 >to work on the ones that will satisfy our objectives.
> >>                 >
> >>                 >Cheers,
> >>                 >Brad
> >>                 >
> >>                 >                 -----Original Message-----
> >>                 >                 From:   ghiasi
> >>[mailto:Ali.Ghiasi@xxxxxxxxxxx]
> >>                 >                 Sent:   Thursday, July 27, 2000 2:17 PM
> >>                 >                 To:     stds-802-3-hssg@xxxxxxxx;
> >>bradley.booth@xxxxxxxxx
> >>                 >                 Cc:     Ali.Ghiasi@xxxxxxxxxxx
> >>                 >                 Subject:        RE: Equalization and
> >>benefits of Parallel
> >>                 >Optics.
> >>                 >
> >>                 >                 Brad
> >>                 >
> >>                 >                 > From: "Booth, Bradley"
> >><bradley.booth@xxxxxxxxx>
> >>                 >                 > To: stds-802-3-hssg@xxxxxxxx
> >>                 >                 > Subject: RE: Equalization and benefits
> >>of Parallel Optics.
> >>                 >                 > Date: Wed, 26 Jul 2000 18:29:56 -0700
> >>                 >                 > MIME-Version: 1.0
> >>                 >                 > X-Resent-To: Multiple Recipients
> >>                 ><stds-802-3-hssg@xxxxxxxxxxxxxxxxxx>
> >>                 >                 > X-Listname: stds-802-3-hssg
> >>                 >                 > X-Info: [Un]Subscribe requests to
> >>                 >majordomo@xxxxxxxxxxxxxxxxxx
> >>                 >                 > X-Moderator-Address:
> >>                 >stds-802-3-hssg-approval@xxxxxxxxxxxxxxxxxx
> >>                 >                 >
> >>                 >                 >
> >>                 >                 > I have one question:
> >>                 >                 >
> >>                 >                 > Which of our distance objectives is
> >>satisfied with
> >>                 >parallel fiber and
> >>                 >                 > parallel optics?
> >>                 >
> >>                 >                 The 100 m data center applications.
> >>                 >                 >
> >>                 >                 > It has been my interpretation that when
> >>we talked about
> >>                 >100m of installed
> >>                 >                 > base of MMF, that we were referring to
> >>the MMF fiber
> >>                 >currently available for
> >>                 >                 > use by 802.3z.  Parallel optics does not
> >>operate over this
> >>                 >installed base.
> >>                 >
> >>                 >                 You are correct parallel optics would not
> >>operate over an
> >>                 >installed two fiber
> >>                 >                 plant.  Parallel optics would loose if you
> >>go in to an
> >>                 >installed fiber base.
> >>                 >                 What I suggested was 100m data center
> >>applications, where
> >>                 >the fiber are not
> >>                 >                 installed in the building wiring.
> >>                 >
> >>                 >                 Data center application are very
> >>significant as stated in
> >>                 >the last meeting
> >>                 >                 about half the total market.  Solutions
> >>significantly lower
> >>                 >cost targeted
> >>                 >                 for sub 100 m is needed, otherwise there
> >>will several
> >>                 >proprietary solutions.
> >>                 >                 Parallel optics is the lowest cost, almost
> >>mature after 3
> >>                 >years, lowest power,
> >>                 >                 and smallest foot print.  Parallel optics
> >>is ideal to get
> >>                 >bandwidth off the
> >>                 >                 edge of your board.
> >>                 >
> >>                 >                 Serial 850 or CWDM 850 can be another
> >>candidate for low cost
> >>                 >data center
> >>                 >                 applications by having cable advantage
> >>over parallell fiber.
> >>                 >But you need
> >>                 >                 to offset fiber advantage against power,
> >>size, cost,
> >>                 >testing, and maturity.
> >>                 >
> >>                 >                 >
> >>                 >                 > Or am I missing the point here?
> >>                 >                 >
> >>                 >                 > Cheers,
> >>                 >                 > Brad
> >>                 >
> >>                 >                 Thanks,
> >>                 >
> >>                 >                 Ali Ghiasi
> >>                 >                 Sun Microsystems
> >>                 >
> >>                 >                 >
> >>                 >
> >>
> >>                 Paul A. Bottorff, Director Switching Architecture
> >>                 Enterprise Solutions Technology Center
> >>                 Nortel Networks, Inc.
> >>                 4401 Great America Parkway
> >>                 Santa Clara, CA 95052-8185
> >>                 Tel: 408 495 3365 Fax: 408 495 1299 ESN: 265 3365
> >>                 email: pbottorf@xxxxxxxxxxxxxxxxxx
> 
> Paul A. Bottorff, Director Switching Architecture
> Enterprise Solutions Technology Center
> Nortel Networks, Inc.
> 4401 Great America Parkway
> Santa Clara, CA 95052-8185
> Tel: 408 495 3365 Fax: 408 495 1299 ESN: 265 3365
> email: pbottorf@xxxxxxxxxxxxxxxxxx
                                   
------------------------------------------------------- 
Richard Taborek Sr.                 Phone: 408-845-6102       
Chief Technology Officer             Cell: 408-832-3957
nSerial Corporation                   Fax: 408-845-6114
2500-5 Augustine Dr.        mailto:rtaborek@xxxxxxxxxxx
Santa Clara, CA 95054            http://www.nSerial.com