Suddenly, my FlightAware stats have gone screw-y again. I use the ranking data to assess the ongoing ‘health’ of my station, but suddenly - like within the last few days - my “Highest Rank in Last 30 Days” field has disappeared completely, and my dynamic ranking that changes during the day, seems to have taken a very distinct downward turn. It gets a little worse overall every day. This is despite nothing changing. I am still recording the same amount of aircraft and same amount of messages as I ever do. The graphs that show this are largely flat. I still remain at the top of the ‘local’ ranking list, which would suggest that the station is working correctly. As far as I can see, I am logged in to my account, and the ‘last feeder check-ins’ are normal. So anyone have any idea what is going on here, and how I at least restore the ‘Last 30 days’ figure, please ?
My Highest ranking in the last 30 days ranking number is never correct. Never has been. I pay no attention to that data line. My station performance remains very consistent but my ranking is constantly changing. I believe it is the changing performance of sites all around the world that drives your ranking and not your specific performance when you are consistent.
Lawrence, I can’t imagine how they are coming up with those numbers. Maybe it has been over engineered and they ended up with an Algorithm that approaches infinity.
Thinking about this a bit more, the “problem“ is that the 30 day ranking is a user ranking, not a site ranking so is of no use at all to compare a site’s performance. A user with a single site will be beaten in the rankings by another user who has 2 sites, that each have a site ranking of 51% of the single site user. Maybe there should be a 30 day site ranking figure too?
Edit, my bad. The dynamic figure is the 30 day site ranking!
The ranking algorithms are sort of weird. I’ve been tracking a particular station of mine since replacing some components after a record-breaking microburst caused water infiltration into my previous antenna. There was significant downtime, and then I had it hooked up again and made a couple of different adjustments at various times. Anyway, I’ve been tracking both rank and rank total (from which the rank is calculated, I think). What’s interesting is that one of them resets at midnight local time. The other resets at midnight UT, hence the jaggies of the two curves don’t coincide. There are jaggies in the first place because the 30 day stats are reset daily, but the current day’s numbers are added throughout the day. That’s why on the dark blue curve, throughout each day, the line goes up until the stat is reset, and on the light blue curve, throughout each day, the line goes down, until the stat is reset. This overall trend of the dark line going up and the light line going down is because I started the graph during INOP, basically, so it took some time to accumulate 30 days of data. I’ve also noticed that sometimes my rank gets worse even though the total count that it’s based on goes up. I conclude from that that everyone else’s numbers are also going up. Maybe that’s because the total number of planes flying is increasing.
OK, and thanks for your reply. Thanks also to all the other good folk that have taken the time and effort to try to explain this. My “30 day ranking” is indeed now back. I have done nothing at all to achieve this. I simply saw your reply and went and had a look, and there it was, after being MIA for 3 days. To be honest, I have never understood quite how the ranking system works. For that reason, I don’t even try to derive any sensible numeric conclusions from it, but however it is derived, i do find it overall useful as a general indicator of the station’s performance. My antenna, for instance, is up quite high and pretty inaccessible from the ground. It is totally exposed to extremes of weather, so is the most likely thing to cause any degradation in station performance. The ‘throughout-the-day’ rolling rank, and the ‘highest in 30 days’ (whatever that actually means) rank give me a very quick visual report of how the station is performing. I do use the other graphs and sets of figures to do a deeper dive every few days, and those seem to do quite well at truly reflecting what is going on. I have had people in the past chiding me to not become obsessed with rankings, but I really think that I don’t. I just want the station performing as well as it can for both my own gratification and the good of the community system that it’s feeding, and if the stats help me do that, then that’s good enough for me.
Anyway, again, thanks to everyone for their input, and my issue seems to be resolved, at least for now
Sorry, I missed your response. And the answer is no. I had never heard of this tool. All of my statistics and statistics analyses are done using the graphs and figures produced by My FlightAware account. Is it something that is easy to set up and can be left running in the background in an open tab all the time ?
Yes there is a very simple install link on Github that is maintained by user Wiedehopf. It’s a great program that provides you lots of real-time and historical information I expect you would fine very useful. I don’t have the current link but I am sure someone can follow up with that link for you.
Note on data loss: When removing or losing power you will lose graph data generated after 23:42 of the previous day. To avoid that issue sudo shutdown now before unplugging the pi. See the section on reducing writes for more detail.
Yeah, sorry. I’ve been really ill all weekend. The wife has had this vicious cold or flu or something and it looks like I’ve got it now …So, this utility runs on the Pi alongside Dump1090 and then its outout data is presented on the same webserver as the FlightAware Data ? Or does all the graph processing get done on the PI as well ? Sorry, I’m well thick when it comes to software, and especially Pi stuff. Also, if it has to run on the Pi, I doubt my current configuration has the capacity.I always have a warning message in Flight Aware that says that the Pi is pretty much at capacity.
Sorry to hear your household has come down sick. It’s my understanding that the Graphs1090 runs locally on the Pi. I run a Rpi 4B+ so I have no capacity issues at all. I originally had a Pi Zero W and for sure that had capacity issues with most everything so I booted that one to the closet. Hope you aren’t suffering from the Pi Zero limitations. I doubt the Graphs1090 uses much resources but the man who wrote it would be able to confirm or deny that. Make sure to get the antibiotics if you need them to keep the Big P out of your chest. Never a good thing.
To answer, no. It has nothing to do with flightaware or it’s associated web or stats pages.
It’s an addon package that runs locally on the capture device. You cant get to it over internet or from anything FlightAware associated. But that reason alone makes monitoring performance better as it takes away the remote data transfering and any filtering or other such criteria imposed.
Its been quite a long time since I put this system together. Its all boxed up in a waterproof enclosure, and not easy to get to. But yes. I seem to recall that it is a Pi Zero. I’ve actually just bought a 4B to start running Satdump on so I can receive and process Meteor weather sats, as they are about to close down the last APT NOAA one. I wonder if you can run two SDRs and two separate applications on a Pi simultaneously
