Network Overview Switch

[wg]

Greetings!

in Network Overview, how is it possible that win7 recognizes the switch
between PC1/PC2 and the gateway?
What is the command/trick/whatsoever to detect an unmanaged switch?

The switch is a simple unmanaged swich.

See http://img441.imageshack.us/img441/2660/networkoverview.png

Thanks for hints,
Wolf

 

[...winston]

It's possible via this approach:
http://en.wikipedia.org/wiki/Link_Layer_Topology_Discovery

The pic you linked to as an example is just a graphic (i.e. topology)


--
....winston
msft mvp mail


"wg" wrote in message
Greetings!

in Network Overview, how is it possible that win7 recognizes the switch
between PC1/PC2 and the gateway?
What is the command/trick/whatsoever to detect an unmanaged switch?

The switch is a simple unmanaged swich.

See http://img441.imageshack.us/img441/2660/networkoverview.png

Thanks for hints,
Wolf

 

[Char Jackson]

"wg" wrote in message
in Network Overview, how is it possible that win7 recognizes the switch
between PC1/PC2 and the gateway?
What is the command/trick/whatsoever to detect an unmanaged switch?

The switch is a simple unmanaged swich.

See http://img441.imageshack.us/img441/2660/networkoverview.png

It's possible via this approach:
http://en.wikipedia.org/wiki/Link_Layer_Topology_Discovery

Expanding a bit on Winston's reply, the LLTD protocol generates some
specific frames of network traffic, then it asks its LLTD Responders
to reply with which of those frames they saw. Responders that _don't_
see specific frames can be assumed to be on the other side of a
switch. As you can see, the switch doesn't actively participate in any
of this discovery and therefore, the fact that the switch is unmanaged
is irrelevant.

See "Section 1.3.2 Topology Discovery Tests" of the LLTD specification
for details, including the following:

"For example, the application using LLTD might do the following. The
application might direct one responder to use a specific LLTD MAC
address and train a switch about the segment to which it is connected
by sending a frame from that MAC address. The application might then
invoke LLTD to request that a second responder send a frame to that
MAC address. Finally, the application could ask a third responder for
the test frames that it saw. If the third responder did not see the
test frame (after multiple such tests to reduce the chance of packet
loss), the application can conclude that the first and third
responders are on different segments. In other words, a switch
separates the segments."

The LLTD spec document can be downloaded here:
<http://msdn.microsoft.com/en-US/windows/hardware/gg463024>

 

[wg]

Expanding a bit on Winston's reply, the LLTD protocol generates some
specific frames of network traffic, then it asks its LLTD Responders
to reply with which of those frames they saw. Responders that _don't_
see specific frames can be assumed to be on the other side of a
switch. As you can see, the switch doesn't actively participate in any
of this discovery and therefore, the fact that the switch is unmanaged
is irrelevant.

See "Section 1.3.2 Topology Discovery Tests" of the LLTD specification
for details, including the following:

"For example, the application using LLTD might do the following. The
application might direct one responder to use a specific LLTD MAC
address and train a switch about the segment to which it is connected
by sending a frame from that MAC address. The application might then
invoke LLTD to request that a second responder send a frame to that
MAC address. Finally, the application could ask a third responder for
the test frames that it saw. If the third responder did not see the
test frame (after multiple such tests to reduce the chance of packet
loss), the application can conclude that the first and third
responders are on different segments. In other words, a switch
separates the segments."

The LLTD spec document can be downloaded here:
<http://msdn.microsoft.com/en-US/windows/hardware/gg463024>

Thanks a lot for the explanation.
Wolf