Piaware mlat issues

hi to all

since one week i have experienced some mlat issues on piaware or at least on flightware webpage, because i see a error saiyng the mlat is not syncronized or location is wrong, the problem goes down afther i restart piaware but normal 1 or two days afther i have same error again, like i said that´s happen only since one week.

it´s really my site problem or flightware web site bug?

thanks in advanced

Did you run the to command to see how much cpu is being used. It could also be a bad dongle… As that was my issue… The dongle was bad… Swapped it and all was fine

I’d also consider try swapping the PSU … if power is low than the effect is usually seen first by the dongle not working properly. If the dongle is on the edge then the MLAT might have issues.

…and it’s an easy thing to do as a test.

(remember if you do want a good USB supply, you can often get them from car accessory shops in a cigarette lighter package to convert 12 or 12-24v down to 5v and with a 3A rating - and you dont need to keep the circuit board inside the funny round case)

I’ve been having similar issues on my Pi2 running PiAware (CPU usage is around 10-15% with a 1/5/15 minute load average of 0.62, 0.56, 0.45) even when changing dongles.

No idea why it shows the warning, but MLAT is still being computed and blue aircraft are being shown (and updated) on my dump1090 web interface.

I’ve also seen some unexpected crashes of dump1090 on one of my Pi2s today. The other Pi2, running the same hardware (I bought two Pi2 kits and two of the same dongles) doesn’t ever have any issues. Go figure.

Other than the odd dump1090 crashes, it’s not really a big deal.

thanks for all answers i think if it was from dongle it’s supposed give me the same issue on FR24 right? but does not, then I will continue to search for solution.

Do you have any USB errors logged?

dmesg -T

Spent another 20.- € today for a new R820T2 dongle, connected to the Pi directly, and still getting the anomaly. :angry:

dmesg -T just reports init stuff from booting, nothing else.

This, slowly, is really going to make me wanting to throw all the ADS-B related stuff into the garbage and convert my Pi2 to a weather station…

That’s no fun.

Out of curiosity, where are you able to buy such dongles at retail shops?

I’m in Switzerland and have never seen them for sale at any of the shops: I just end up getting them from Amazon in the US (even with shipping it’s cheaper than other Swiss or German vendors I’ve seen online).

No, I’ve ordered it online and received it today, so it’s kind of like I’ve spent the money today. :wink:

Still wondering what it is. WIFI/Internet connection latency issue? Server issue? MLAT implementation issue?

It also occurs at other peoples Pi’s, so it’s not only sporadic and someone should have a closer look at it…

Ignorance is bliss?

The anomaly reports are not reporting anything new; they’re just giving you visibility on the quality of information as the mlat server sees it.
The instability of some dongles seems to be simply a hardware thing as far as I can tell (unless someone can point me at another reason why the dongles would randomly drop 10-20us of samples)

also note that if it’s a transient or intermittent problem, the server will happily start using the information again once the quality improves (on a timeframe of minutes)

Once I get a new piaware release sorted out you can look at the piaware logs for more of a history of the mlat synchronization quality over time.

i ill thank you for people stoping saying the problem its from dongle site because is not, really tahnk you.

I think I now can tell that this issue is connected to temperature. Since autumn arrived and the weather cooled down here, this did not show up again. It mainly was present during the hot weather period.

So, the question is, which part is causing the effect?

I would think of the the RTL2832 or the RasPi USB/Ethernet-controller, but this are just first guesses.

Maybe someone shoud do some old-school hardware debugging, using ice-spray an a blow-dryer? :wink:

@obj: Can you confirm that this effect is getting more rarely, in areas where the weather starts to cool down?