[Gllug] How to prove a computer is overheating.

[Chris Bell]

On Sun 16 Jan, Peter Childs wrote:
> 
> Chris Bell wrote:
> 

> >   I asume the ethernet cables are OK, with no data corruption.
> >
> >  
> >
>     Can I not answer that question as the network is a bad subject,
> 
>     I mostly wired it. If you know any way of testing it I would love to 
> hear it.
> 
>     Its mostly 100Mbts (Switches) with CAT5 cables the only bit thats a 
> bit anoying is the 10Mbts subnet running over good old Coax.
> 
> Peter Childs

   I only have a two-part cable tester which can check whether Cat5 pin to
pin connections are correct and good, with no shorts or open circuits, but
not whether the correct wires have been used in the correct order. Most
problems are associated with cables connected using an incorrect sequence of
wires, because the layout is not obvious. The wire order can usually be
checked using a magnifying glass, and some idea of the data error rate by
typing

ifconfig <return>

   Coax has gained a poor reputation, possibly because it is not always
mechanically robust, it suffers from every suspect connection along the way,
and must be correctly terminated at both ends of the run.

Standard ethernet connections 1 to 1    Cross wired cables are connected
         ___________________ 
         |                 |
W/O  ----|                -| 1          1  W/O  3  (W/G)
O/W  ----|Cable  Contacts -| 2          2  O/W  6  (G/W)
W/Gn  ---|............... -| 3          3  W/G  1  (W/O)
Bu/W  ---|              . -| 4          
W/Bu  ---|............... -| 5          
Gn/W  ---|                -| 6          6  G/W  2  (O/W)
W/Bn  ---|(Latch below)   -| 7          
Bn/W  ---|                -| 8          
         |_________________|

Pairs DO cross over in the middle!
Only two of the pairs are normally used, the others could be used for
telephone or a second data set using splitter adapters.

-- 
Chris Bell

-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug