|
- Link Tracing using the Parascope 2000
The Parascope 2000 is a WAN analyser with v.24, x.21 and v.35 interfaces
which allows you to analyse various protocol types over a variety
of physical media. The Parascope 2000 can be useful if you wish to
analyse a Link network containing Datacryptor®s. One important fact to understand
about this process is that this kind of analysis does not show anything
recognisable as user data - because the Datacryptor® Link operates at the
link layer (ones and zeroes) and does not have any framing structure
in the way that IP and Frame Relay do it is not possible to see any
structure to the data.
However, the Parascope 2000 does give you information on how much
data is transmitted from DTE/DCE, throughput etc. One particular
case where it is useful is if you have a Datacryptor® and you suspect that
one or more of the interfaces is faulty you can test whether any
data is being transmitted or not. The LED display also shows the
status of the signals on the line (CTS, DTR etc.).
The example used below uses x.21.
Click
on picture thumbnail to view diagram
As can be seen in the diagram above, the Parascope 2000 sits in-between
the Datacryptor® and the null modem and transparently captures the data.
If you wish to pass real data over the link you would need to setup
a device on the host side of each Datacryptor® (see the section on setup
with Cisco routers ). You can still see traffic generated on the
link by using the diagnostic ping or by forcing key exchanges.
- BERT testing using the Parascope 2000
BERT (Bit Error Rate Test) is a way of passing data through device
to check for the existence of errors on the line - it is a good way
of testing a Datacryptor® if a customer reports errors on the line or a faulty
interface. The Parascope 2000 analyser is able to do BERT testing.
The network is setup as below:
Click
on picture thumbnail to view diagram
Note the x.21 loopback plug (See appendix for pin outs) on the host
port of the far Datacryptor® - this allows data to be looped back through
the system. An encrypted tunnel is created between the two units.
In this setup we are passing data through both units - the Parascope
send a data stream which goes through the system and is then looped
back at the end. On its return to the Parascope the data stream is
checked for errors and a line grade (the quality of the line is reported).
To setup the Parascope 2000 to perform the BERT test follow the instructions
below.
- On the Quick Start window select the x.21 box:
Click
on picture thumbnail to view diagram
- Then click on the top Setup button to the right of the Line Interface
selection:
Click
on picture thumbnail to view diagram
- Click on the Sync tab and make sure that the clock source is set
to external. Click on OK to exit.
- Click on the BERT radio button and then click on the bottom Setup button to the right.
Click
on picture thumbnail to view diagram
Ensure that the clock source is set to external (we do not need
to provide a clock as the null modem does this for us). Click on
OK to continue.
Click
on picture thumbnail to view diagram
- When the BERT tests starts you will see the screen below:
Click
on picture thumbnail to view diagram
This shows that the BERT test is "out of lock" as
indicated in the top right corner. This means that the data is being
sent but
not being returned. This should go after a few seconds - if it does
not then there is a problem with the setup (usually cabling). Note
the clock frequency (347911Hz) - this indicates the line speed as
specified on the null modem. Once the BERT test goes into lock you
should see this screen:
Click
on picture thumbnail to view diagram
The diagram above shows a good quality line with no errors - note
the % Error Free Sec showing 100.00% and a line grade of EXCELLENT.
If you see errors they will be displayed in the bottom left corner
in red and the line grade and % error free sec will degrade. It is
possible to insert errors using the WinXL application - you will
also see errors if you are peer managing the remote unit whilst running
the BERT test.
You can leave the BERT test running for as long as necessary - it
will soon become evident whether there are errors being generated.
This kind of setup tests both units so you can't really see which
unit is causing the problem if there are errors being generated.
If you have a faulty unit the best way to approach this is to start
the test of with two known good link units and get a steady line
grade and then swap one of the good units for the faulty one and
restart the test.
|
)
)
)
)
)
)
)
)
