change in network speed...

change in network speed...

Post by Rick S » Sun, 01 Jun 2003 04:29:31



My father has a very, very small Ethernet network in his house.
2 or 3 computers.

He uses it to back-up data from one PC to the other. Usually
this back-up would take 15 seconds. But a few days ago he
was "dusting" (i.e. cleaning up) and he might have moved the
Ethernet cables a tiny bit.

Now the back-up still works, but it takes 2 minutes.

As the cables are store-bought, not home-made, my first
guess would not be them, but maybe.

But if the cable were now passing close by an electrical
power cable, would that cause trouble ?  I would guess that
thay are "shielded", but maybe that's the problem, just the same ?

I assume what is happening is lost data, but TCP/IP just keeps
re-trying until it gets everything right, and after 2 minutes,
everything is.

Any other ideas ?  Thanks very much.

 
 
 

change in network speed...

Post by Fester Ranci » Sun, 01 Jun 2003 05:10:44


Diagnosis sounds pretty good... moving the cables could cause cable damage
but the plugs are more likely to be damaged than the cable, however it's not
unheard of. Doing stuff like running cables where people are likely to stand
on them, or putting too sharp a bend in a cable is not good. If I remember
correctly the minimum bend radius is about 38mm.

BTW the U in UTP (as in UTP patch cable) stands for "Unshielded"; you are
unlikely to have bought a shielded patch cord (FTP, where the "F" stands for
"fully shielded").

FR
--
There is nothing more frightening than ignorance in action. - Goethe


Quote:> My father has a very, very small Ethernet network in his house.
> 2 or 3 computers.

> He uses it to back-up data from one PC to the other. Usually
> this back-up would take 15 seconds. But a few days ago he
> was "dusting" (i.e. cleaning up) and he might have moved the
> Ethernet cables a tiny bit.

> Now the back-up still works, but it takes 2 minutes.

> As the cables are store-bought, not home-made, my first
> guess would not be them, but maybe.

> But if the cable were now passing close by an electrical
> power cable, would that cause trouble ?  I would guess that
> thay are "shielded", but maybe that's the problem, just the same ?

> I assume what is happening is lost data, but TCP/IP just keeps
> re-trying until it gets everything right, and after 2 minutes,
> everything is.

> Any other ideas ?  Thanks very much.


 
 
 

change in network speed...

Post by Glen Herrmannsfeld » Sun, 01 Jun 2003 06:00:38



Quote:> My father has a very, very small Ethernet network in his house.
> 2 or 3 computers.

> He uses it to back-up data from one PC to the other. Usually
> this back-up would take 15 seconds. But a few days ago he
> was "dusting" (i.e. cleaning up) and he might have moved the
> Ethernet cables a tiny bit.

> Now the back-up still works, but it takes 2 minutes.

> As the cables are store-bought, not home-made, my first
> guess would not be them, but maybe.

> But if the cable were now passing close by an electrical
> power cable, would that cause trouble ?  I would guess that
> thay are "shielded", but maybe that's the problem, just the same ?

> I assume what is happening is lost data, but TCP/IP just keeps
> re-trying until it gets everything right, and after 2 minutes,
> everything is.

Running too close to other cables, other than those powering arc welders
(that has been asked here) doesn't tend to be a problem.

I have known connectors to make just enough contact to sometimes work, with
symptoms like that.   Check the link lights for all connections.

-- glen

 
 
 

change in network speed...

Post by Paul T Wan » Sun, 01 Jun 2003 11:02:48



Quote:> Diagnosis sounds pretty good... moving the cables could cause cable damage
> but the plugs are more likely to be damaged than the cable, however it's
not
> unheard of. Doing stuff like running cables where people are likely to
stand
> on them, or putting too sharp a bend in a cable is not good. If I remember
> correctly the minimum bend radius is about 38mm.

> BTW the U in UTP (as in UTP patch cable) stands for "Unshielded"; you are
> unlikely to have bought a shielded patch cord (FTP, where the "F" stands
for
> "fully shielded").

Do you mean when something is pressing on the cable, the chance of getting
errors actually increase?

What kind of interference can affect network cables?

 
 
 

change in network speed...

Post by Fester Ranci » Mon, 02 Jun 2003 03:47:03






> > Diagnosis sounds pretty good... moving the cables could cause cable
damage
> > but the plugs are more likely to be damaged than the cable, however it's
> not
> > unheard of. Doing stuff like running cables where people are likely to
> stand
> > on them, or putting too sharp a bend in a cable is not good. If I
remember
> > correctly the minimum bend radius is about 38mm.

> > BTW the U in UTP (as in UTP patch cable) stands for "Unshielded"; you
are
> > unlikely to have bought a shielded patch cord (FTP, where the "F" stands
> for
> > "fully shielded").

> Do you mean when something is pressing on the cable, the chance of getting
> errors actually increase?

Yes!

Twisted pair cables as used for data networks are constructed so that each
pair is balanced; i.e. the pair is twisted round each other ata specific
rate of twist. The four pairs are twisted at different rates so that the
can't interfere with each other. The four twisted pairs are then twisted
together, again to reduce the chance of causing or recieving interference.

When a twisted par cable is laid straight the cable is perfectly balanced.
However, should the cable be bent, or if pressure is applied to the cable
then it can push the twists out of balance and increase the chance of some
sort of interference occuring. Likewise if a cable is subject to too much
tension (like when pulling it off a drum), it can by damaged just by pulling
on it.

This is why when data cabling is professionally installed it should be laid
using a route with the minimum number of changes of direction.

Quote:> What kind of interference can affect network cables?

Running data cables alongside power cables is generally thought to be bad
practice these days, (with the network speed going up the likelihood of a
small amount of interference causing network problems increases). There is a
recent European data installation standard for the use of 3-compartment
trunking which states that the data & power cables should be in opposite
compartments with voice cabling using the centre compartment; i.e. if they
have to run parallel then have them as far apart as possible.

FR

 
 
 

1. Changing the modem speed from port speed to actual conneting speed


You saw Art's answer. Bet you now want to know "How?"

That depends on the model of the modem - different modems use
different commands.  It _also_ depends on if the modem configuration
is actually the problem - it may not be.

DUN will display the DTE (port) speed under two conditions:
1 - the modem returns the CONNECT message with the port speed (modem
config problem)
2 - the modem is installed with the wrong .INF file - not a modem
config problem.

In this latter case DUN tries to match what the modem sends with it's
connect message with data provided from the .INF, and if it can't get
a match, it reports the port speed.

If you're _positive_ the correct .INF file was used to install the
modem:

Try adding W2 to Extra Settings.  If that doesn't work, try adding
S95=1.

And if that works, ponder why the idiots who wrote the .INF file
didn't properly configure the modem in the first place.

2. Use of RX26 (DVA0:) on AXP 4000/233???

3. Change displayed connect speed from port speed to modem speed?

4. Acorn User Web site

5. Will change bandwidth on serial port change actual speed?

6. AHHHH!! I need a good 28.8!!

7. speed change in network

8. JOBS OFFERED(Contractor or W2)

9. INS (International Network Services)

10. New network - dual speed hub, dual speed switch, or two hubs

11. Modem init string help and ways to speed up your modem connection dial up network networking

12. changing dhcp lease time and speed issue

13. Changing Speed on AUX 0