[En-Nut-Discussion] A different problem with ARP...

[Michael Jones]

Hello,

I've just stumbled into a new (?) problem which seems to involve ARP. 

What I see is that ARP requests are sent from the Nut/OS and that the Nut/OS
gets sent an answer be the corresponding device. Yet, Nut/OS does not stop
re-requesting the same information over and over again. Any currently open
TCP/IP connection between the two devices eventually disconnect...

The other symptom is that a device will do an ARP request over and over
again but the TCP/IP connection is actually still working (which is a
programmatic contradiction with the ARP request) and when I ping the device
from the other end the ARP stops. 

Unluckily these symptoms are not reproducible "in the lab"... 

Any ideas? Has anybody observed similar? Does anybody know how to trigger
this behavior?


Cu,
Michael