Radio RRC Stopping During Contest

[K3TN]

Been running K3 Twin in the ARRL DX CW, two different people taking turns connecting to the remote K3. We made over 3200 QSOs by early this morning when all of a sudden, in the middle of a run on 20m, the radio stopped responding (I have a K3?mini, the other guy is using his K3 in terminal mode.)

I could still hear rx audio, and I would hear sidetone when I hit the key, but the TX light wouldn't light on the K3 Mini, and there was no blanking of the RX audio when it was "sending" So, audio continued, data comms did not. Turned the K3 Mini off and on again - and got the busy signal and no connection whatsoever to the remote K3.

Found that the radio RRC could not be reached via browser. Pinging the remote router, showed all is fine there. Doing a remote VNC session to a PC on the radio LAN, still couldn't browse the RRC. The Microbit Setup Manager Find Devices command would find it, but it would launch a browser, and get no response. Pinging the RRC internal address from the radio LAN got no response.

A few hours later it started working again. Got back on and starting running like crazy on 10M, went for an hour or so and same failure mode.

Big snowstorm here means no one can get to the radio site, trying to figure this out remotely. Power issue? RFI? Software bug?

We are running the 10 Feb 2.85 version that has been stable.

Waiting to see if it comes back again, or the snow melts - whichever comes first...

John K3TN

[K3TN]

A bit more data: the Radio RRC came back on its own after 2-3 hours,  and connecting to it via browser, the status did not show any downtime and there was no saved error message. It seems like the RRC kept running but the IP stack just wouldn't respond.

It worked for a few hours, then died again for several more hours. Then came back very briefly and stayed dead for over 8 hours - and then came back and seems to have stayed back up.

One possible common denominator: It only died when I was connected, I'm running the same 2.85 10 Feb firmware that the Radio RRC has. N1RM was also connecting at times and it never died when he was using it and it turns out he was running the 2.84 firmware load. He was never hitting it as hard as I was - I was almost always running at 120+/hour rates with lots of Fkey hitting and Winkey commands flying over the connection, while Rick was mostly doing S&P.

What failure mode would cause RRC to be unresponsive to SIP or HTTP connection or more than 1 ping, but firmware would not report downtime?

[Jan (Microbit)]

What failure mode would cause RRC to be unresponsive to SIP or HTTP connection or more than 1 ping, but firmware would not report downtime?

What you describe sounds so strange that I would call it almost "impossible", but experience tells that there are new "impossible" things happen every day, so I am not saying that. What is strangest is that it was "partial alive" by responding to "Find Devices" commands(which uses UDP), and yet it did not reply to pings. Also strange is that it did not restart, nor showed and error messages.

Did you try to restart the RRC while it was workable?

[K3TN]

Yes, when I came back I did a remote restart but it went away again later. Right now I'm reloading the last 2.85 firmware to see if that changes anything.