Conference netcad::hub_mgnt

    The Ethernet spec states 2800 M is the maximum so that shouldn't be
    your problem.  The distance should be based on the loss budget.  But I
    don't what the transmit and receive specs are however on the DB 90FL or
    what the fiber loss is at 850 nm.
    
    In a similar fashion we can run 60 KM SMF between FDDI stations assuming
    excellent splices and low loss fiber.
    
    dave

    The 2800 meter rule does not mean 2800 meters of fiber, it means 2800
    meters between any 2 stations, which is comprised of 1500 meters of
    thickwire, 1000 meters of MM fiber and 300 meters of transceiver cable.
    It assumes the following
    1 - 3 500 meter segments of thickwire
    2 - 2 500 meter segments of MM fiber
    3 - 6 transceiver cables no longer than 50 meters, 4 for the repeaters
    and 1 each for the 2 stations.
    
    This is right out of Ethernet V2.0 specification.
    
    With that in mind use the FL spec as your rule and if the customer
    chooses to violate that rule (and it will work if you have VERY low
    loss on the 2700 meter fiber segment) it will not be supported by
    Digital. Customers violate the configuration all the time and you MUST
    state to them it is a violation and Digital will not support it
    otherwise we just bought the farm!

    Also see notes file KALI::OPEN_DECCONNECT
    Entry 102 and replies, "Ethernet rules and NETdesigner"
    
    Dave
    

	Okay, so it looks like 2000M is the max for these bridges.

	1) What can he substitute to accommodate the 2700M requirement?
	The background is that he took an order based upon using
	the DEBAM-RC.  When he went to order the DEBAM's they were 
	end of lifed.

	2) Anyone have any experience using Cabletron's FOT-F24 or
	Chipcom's 10Base-FB tranceivers?

	

    The way I understand it is that if you have overcome the extra loss
    problem, when going 2700 meters, there is still the problem of pulse-
    widening, which occurs due to the variety of signal travel time caused by
    the "multiple modes" of propagation. Signals arrive at the far end after 
    having travelled inside the fiber for slightly differing amounts of time, 
    and when these signals are all combined at the receiving end, the
    resultant signal may not look as "clean" as it was at the transmitting end,
    and that can cause excessive errors.
    
    That's the reason "single-mode" fiber goes much farther, since there
    is no way for light rays to reflect at diferent angles inside that very
    fine fiber.
    
    As for the extra loss there should be no problem, if patch-panels and
    other lossy connections and splices are avoided. The 2000m limit allows
    for those, and the extra db's so gained can now be used for extra fiber. 
    
    In retrospect, DEC had a product that communicated over 3000m, using
    the same 62.5/125 fiber and the same 850nm wavelength. Although the
    standard was not 10baseFL, the signals transmitted over such a link
    were the same 10Mb/s Ethernet signals as today, so it is my feeling
    that if 3000 meters was OK then, it should be so today, even though it
    is NOT SUPPORTED BY DEC, and of course, the suggestion in a previous 
    answer, to obtain a WRITTEN SIGNOFF from the customer, makes very much 
    sense.
    
    
    
    
    
    

Let me say it again, a different way.  

The customer presented a problem to us for which we quoted a *supported* 
configuration; ie the DEBAM-RC aka LAN Bridge 200 (3 km version).  Between 
the time that the customer gave us the PO and we ordered the equipment, the 
DEBAM-RC was end-of-lifed.  

This is *not* the customer's problem; it is Digital's!

This situation was elevated through the appropriate NPB sales chain and no 
*reasonable* solution was offered.  

It is now Digital's (Network Services) obligation to provide a comparable 
*supported* solution. 

With that in mind, can anyone think of a supported Digital solution?  If not,
we will need to look at 3rd party solutions.

Before anyone says it, the DEBAM-RF is not a viable solution due to the much 
higher (3X) cost.

    .re .2
    
    >> 1 - 3 500 meter segments of thickwire
    >> 2 - 2 500 meter segments of MM fiber
    >> 3 - 6 transceiver cables no longer than 50 meters, 4 for the repeaters
    >> and 1 each for the 2 stations.
    
    >> This is right out of Ethernet V2.0 specification.
    
    Yea, but that was before 10BaseFL.  It doesn't matter what is between
    the 2 end stations so long as the length of the individual segment
    doesn't exceed its maximum length, ie. 500M for thickwire.  It all
    works fine so long as the round trip delay doesn't exceed 51.x uS.  I
    can never rememberthe number exactly.
    
    >> With that in mind use the FL spec as your rule and if the customer
    >> chooses to violate that rule (and it will work if you have VERY low
    
    >> state to them it is a violation and Digital will not support it
    
    Agreed.
    
    dave

	Last call,

	Does anyone have a Digital solution to this problem?

    The DEIWX-FA RemoteConnect device, connects two Ethernet LANs up to 
    5 miles @ 2Mb/s throughput.
    
    
    Barry

    Our agreement with Persoft expried back in May.  We are no longer able
    to sell the directional product.  Persoft continues to sell an updated
    version.  The bridging software we were selling was V1, they are
    selling at least V3 now.  Just a little behind.  There also seems
    little interest in doing more that 800 feet right now within Digital. 
    Maybe that will change in the future but now for at least a year.
    
    dave