Followup to reg at lti's "bizarre situation"

[David Talmage]

Summary: Unsatisfactory NFS performance through a BICC 1125 Multiport
Repeater.

A while ago Chris Wilson, munnari!adl.austek.oz.au!cwilson at uunet.UU.NET,
asked us about problems with Suns and ShineNet ethernet repeaters.  His
problem description was so close to the one we had then at lti that I
wrote to him to learn more.  Below is Chris's reply:

>I now have some more information on this. The ShineNet repeaters are
>based on BICC repeaters and the problem is common to all of them. The
>cause is the different crystals controlling the timing in the
>repeaters and the ethernet interfaces on the Suns. Both are in spec
>but at opposite ends of the allowable range. Hence the problem. I have
>information from both Sun Australia and my local supplier of the
>ShineNet this is the problem. Some guy in Holland also had the same
>problem and was also told it was the crystals.

>BICC have been talking to Sun about this and have an alternative
>crystal set available for their repeaters (according to Sun) which
>fixes the problem. ShineNet are preparing a crystal set at this very
>moment and it should be available soon. So I have the same diagnosis
>and solution from 3 sources. I suggest you contact whoever supplied
>your BICC and ask for a replacement crystal. On the ShineNet repeaters
>the crystals are alleged to be field (user?) replaceable,


>Chris Wilson
>Austek Microsystems Pty. Ltd.
>Technology Park 
>Adelaide, SA 5095


One telephone call to BICC later, I disovered that what Chris said is true.
BICC sent a loaner 1125 for me to use while mine is repaired under
warranty.  Sho' 'nuf, when I put in the new 1125, network performance
improved tremendously!


David Talmage

David W. Talmage, Systems Programmer		...!{buita,bbn}!lti!talmage
Language Technology, Inc.			talmage%lti.uucp at bu-it.bu.edu
27 Congress St., Salem, MA 01970		(508) 741-1507