I tested finally a TNC link between 2 computers running CT9.54.
The setup was as written on http://www.k1ea.com/tnc_network.htm
I used 2 TNCs: MFJ-1280 and PK-232 connected in VHF@1200b/s
At one end a serial port was connected to cluster.
The link was 90% of time available. The rest of time was spent to
restore connection. If no previous experience, was difficult to restore
connection during panic situations...
I am concerned about 2 issues:
1. After the last step to configure the TNC link, a new step must be
added: quit the program and enter again. Otherwise, connection is not
good and "Network bad checksum" message appears.
2. For some unknown reason, the link is stable only for a limited amount
of time (30'...3h). Then it stops to provide data for CT applications,
"Network buffer overrun" appears and a new connection is required. TNC
start/stop is mandatory. I belive that cause is the cluster traffic. I
filtered the traffic and I had less disconnections.
If anyone have ideas how to avoid these problems in future, please let
me know. I am sure there are not CT bugs, but user mistakes...
73s de Ciprian YO3FWC
|