[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference 7.286::fddi

Title:FDDI - The Next Generation
Moderator:NETCAD::STEFANI
Created:Thu Apr 27 1989
Last Modified:Thu Jun 05 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:2259
Total number of notes:8590

2019.0. "LEM errors/LEM rejects" by 49575::PAIVA (Hawkeye - Network Support @GEO) Fri Apr 19 1996 12:09

    Hi!
    
    A customer is seeing LEM errors (3'000 in 25 days) and LEM Rejects (85 in
    the same time) on one port of the first DECconcentrator 900TH. On the other
    900TH 80 LEM errors and no LEM Rejects during the same time. They are
    on a hub 900. The remaining ports (one B and one A) are connected through
    the hub's backplane. The ring is as follows:
    
          900TH-----900TH
           �|         |
            |         |
          GS/FDDI-----
    
    However on the oppposite side of the fibers, there are no errors on
    the GIGAswitch/FDDI's ports.
    
    This is not a firmware issue. All are up to date. The attenuations have
    been measured on the fibers and are OK, although we suspect one fiber.
    
    The question are:
    What do LEM Rejects represent? 
    The number of times the link has been rejected thus making the port wrap
    from the primary ring to the secondary ? this doen't seem possible if the
    GS/FDDI doesn't have LEM Rejects. 
    Or could it be that it is always the 900TH that DETECTS the link error and
    then wraps (before the GS/FDDI)? 
    this doesn't sound reasonable either.
    
    One step will be to see what happens if we replace the PMD.
    
    Any advice will be appreciated. 
    
    Pedro PAIVA
    
T.RTitleUserPersonal
Name
DateLines
2019.1LEM Reject = Many link errorsNPSS::KIRKFri Apr 19 1996 15:5048

  	The short answer is that an LEM reject represents a bad link -
     	too many errors.  This can occur if the power into the receiver
     	is too weak, or for a number of other reasons.  Suspecting either
     	the fiber or the PMD are reasonable things to do. Cleaning all
     	fiber connectors is also a good idea.

     	You said that all the attenuations are correct, but that one fiber 
     	is suspect.  Can you provide the details?  I would need fiber
     	lengths and attenuations

     	If you can, try reversing the fiber pair at both ends of the link.
     	If the fiber is really bad, the errors should be noticed at the
     	Gigaswitch.  If the 900TH PMD is bad, then the errors will still be
     	reported at the 900TH.

     	If the fiber is the problem, but is only marginally bad, it may
     	be that the receiver at the GIGAswitch may be better than at the
     	900TH, in which case the rate of errors may decrease.

     	In any case, I would advise use of known good fiber links.     

     Please provide the fiber details, and see if the problem follows
     the PMD, or the fiber.


     Regards,

     	Dick Kirk
     	Network Product Support

     

     Some additional detail on LEM Rejects:

	LEM Rejects can occur if the LerEstimate drops to the LerAlarm 
     level. When a link is established after the FDDI LCT (Link Confidence 
     Test), the LerEstimate is set to 15. The LerEstimate only changes if 
     link errors are detected. It is a measure of the bit error ratio of 
     the link.  Our FDDI links have monitors (LEM) with alarms. If the 
     LerEstimate gets down to 8 (roughly a BER of 1e-08), then the firmware 
     declares the link to be 'broken' since it does not meet the required 
     FDDI bit error ratio. So the connection is rejected by the Link Error 
     Monitor (LEM).

    
    
2019.2Thanks49575::PAIVAHawkeye - Network Support @GEOTue Apr 30 1996 08:257
    Dick,
    
      Thank you very much for your reply. I'm having trouble getting in
    touch with the customer, reason why I didn't post the fiber figures (I
    don't have them). Will advise when I will have more information.
    
    Pedro