Dump 1090 appears to crash with too many aircraft

[DuncanLowe]

Hi. It is cable connected, but being a Pi zero that’s via a USB ethernet adaptor that the SDR is also plugged into.

As I couldn’t get any connection, I wondered if resetting just the dongle rather than the whole device would help. I did and I could see that the Pi was doing something when plugged back in (green light flashing) and I can now SSH. So here are the logs:

pi@piaware : ~ $ sudo systemctl --no-pager status dump1090-fa

● dump1090-fa.service - dump1090 ADS-B receiver (FlightAware customization)

Loaded: loaded (/lib/systemd/system/dump1090-fa.service; enabled; vendor preset: enabled)

Active: active (running) since Fri 2021-06-25 09:17:41 UTC; 19min ago

Docs: PiAware - ADS-B and MLAT Receiver - FlightAware

Main PID: 427 (dump1090-fa)

Memory: 5.5M

CGroup: /system.slice/dump1090-fa.service

└─427 /usr/bin/dump1090-fa --device-index 0 --gain -10 --ppm 0 --max…

Jun 25 09:17:41 piaware systemd[1]: Started dump1090 ADS-B receiver (Flight…on).

Jun 25 09:17:41 piaware dump1090-fa[427]: Fri Jun 25 09:17:41 2021 UTC dum… up.

Jun 25 09:17:43 piaware dump1090-fa[427]: rtlsdr: using device #0: Generic …000)

Jun 25 09:17:43 piaware dump1090-fa[427]: Detached kernel driver

Jun 25 09:17:43 piaware dump1090-fa[427]: Found Rafael Micro R820T tuner

Jun 25 09:17:43 piaware dump1090-fa[427]: rtlsdr: enabling tuner AGC

Jun 25 09:17:43 piaware dump1090-fa[427]: Allocating 4 zero-copy buffers

Hint: Some lines were ellipsized, use -l to show in full.

pi@piaware : ~ $ sudo journalctl --no-pager -n50 -u dump1090-fa

– Logs begin at Fri 2021-06-25 09:17:01 UTC, end at Fri 2021-06-25 09:42:41 UTC. –

Jun 25 09:17:41 piaware systemd[1]: Started dump1090 ADS-B receiver (FlightAware customization).

Jun 25 09:17:41 piaware dump1090-fa[427]: Fri Jun 25 09:17:41 2021 UTC dump1090-fa 5.0 starting up.

Jun 25 09:17:43 piaware dump1090-fa[427]: rtlsdr: using device #0: Generic RTL2832U (Realtek, RTL2832U, SN 00001000)

Jun 25 09:17:43 piaware dump1090-fa[427]: Detached kernel driver

Jun 25 09:17:43 piaware dump1090-fa[427]: Found Rafael Micro R820T tuner

Jun 25 09:17:43 piaware dump1090-fa[427]: rtlsdr: enabling tuner AGC

Jun 25 09:17:43 piaware dump1090-fa[427]: Allocating 4 zero-copy buffers

Also, without any other intervention, the open page showing the error message is now showing aircraft again. So my guess is that the problem lies with the USB / ethernet adaptor!

[tomvdhorst]

I think the usb/ethernet adaptor might indeed be the cause of your problem.
If you have a different Pi at hand I would recommend to try the setup with that one in order to see if it will function correctly.
Normally you should be able to swap the SD card over and it should boot up and start running.
Testing it with a wired connection would give you the best results (without the adapter then).

[wiedehopf]

Run this … maybe it fixes your issue:
Raspbian dropping off the network potential workaround · wiedehopf/adsb-wiki Wiki · GitHub

[DuncanLowe]

OK, thanks, have done that. I’ll keep an eye on it and try it in the other location. It makes me wonder if it’s the network situation in the two locations that makes the difference. The unmanaged switch in the two locations will certainly have different traffic levels, but I thought the fact they are switches would mean the PiAware itself would see only traffic meant for it. Alternatively, it could simply be that the adaptor isn’t up to it, and as suggested swapping the SD into anothe RPi with it’s own network port might be educational as to the problem.

[DuncanLowe]

OK, an update again. I’m getting the same AJAX call message when looking at the web interface. But this time, I guess because the auto restart of the network interface has been done (the system uptime is less than an hour), I can still ‘see’ the RPi. I have two windows open, one logged in viewing the map (now with no aircraft and the error message), and a second not logged in showing basic stats. The latter is showing that the 1090 radio isn’t working (screenshot attached). Given it’s plugged into the same adaptor where the network adaptor is (combined USB and ethernet), I’m getting increasingly suspicious that it’s the root cause.

[tomvdhorst]

I would swap devices to see if that improves the situation. I suspect your combined device is overloading the usb bus of the Pi zero.

[DuncanLowe]

Doing some searching, it seems that not all of these adaptors are created equal. So have ordered a replacement from Pimoroni, that gets decent reviews, some of which actually mention how many have a habit of falling over regularly.

[DuncanLowe]

New adaptor installed, I’ll keep the thread updated. Have to say I’m impressed with Pimoroni, new adaptor through my letterbox in less than 24 hrs.

[DuncanLowe]

Well thanks to everyone for the help and hints.The new adaptor worked flawlessly on the poor antenna, where in recent days the original had started falling over quite often, again seeming to co-incide with lots of traffic.We get a lot of traffic into BHX for example, but also there’s a flying club nearby and we seem to be a mian route for various aircraft (helicopters and light aircraft) maybe because visually it’s easy to follow the Trent Valley? That meant it had been quite busy recently.

So right back to the start of the question. Connecting to the better antenna which brings in extra traffic. Last time with the old adaptor, it wouldn’t stay up for more than a couple of minutes before behaving as described above. It’s now done two hours and I’m seeing way more traffic than normal. So I’m 90% confident the problem was the adaptor. Whether the traffic over the USB or the power consumption because of that traffic I don’t know, but this new adaptor is behaving so far.

Regarding antenna gain, and the balance of seeing distant aircraft, with avoiding overloading the front end, I’ve had a number of aircraft today (from the mentioned local flying school) almost completely overhead and below 2000 ft. There might have been a few missed messages, but certainly no obvious radio front end overload.

I’ll keep monitoring of course, and feed back but maybe this will also help anyone seeing a similar problem. Bottom line it seems a Pi Zero is perfectly capable, but Ethernet / USB hubs are NOT created equal.

[tomvdhorst]

Good to see you have the problem at it’s tail, now just wait a few days to see if it is behaving as expected.

Pi Zero can indeed do the job als long as all the connected equipment is in order as well

[DuncanLowe]

Just to update everyone who helped. The new USB / ethernet adaptor has been rock solid. With the new antenna I’m reporting three times the aircraft (around 400 per day to more like 1200 per day), and the obvious bias to the south has belanced up.Well a rubbish antenna stuck inside a window was never going to be the best. Not only that, but have been able to feed another aggregator without any performance issues.

Now have to decide what to do with the spare Pi I was going to use for the other aggregator. There WILL be something useful.

Again, thanks for the support!