Confused with ADS-B range

My ground plate seems to be working well, the improved version added about 10-11% more detection.
Im running at 46.9 and only have 1.4% strong messages…
Changing to AGC and it perform far worse… I need to look at changing it on a schedule.

But on a another topic i think there might be a leak in one of the processes…Or could I have logging to RAM and it not being committed to Disk?
Today i had a crash, and here is what i saw:

Memory has been growing steadily since this Pi was built.:

After the crash it fell considerably
There was a flurry or activity just before it fell over.

It is the nature of Linux, to use the existing RAM efficiently. Not comparable to e.g. Windows.

My Notebook has 32GB RAM and almost 10% of it are used as buffer just about an hour after restart and usage.

The longer the device is running, the more memory is used. Usually not a bad thing.
But your device has a large increase of “used”, this is not normal. You should check with top/htop which process eats the memory over time.

Dave, Do you have any other processes running on your Pi? For most of us, the Green section is between 150 and 250 Megabytes, Blue is a thin like as is yours, and yellow around 300 Megabytes. Not specific at all, just general info. There is definitely a memory leak in some aspect of your system, and you may find the cause in the system logs. Using Putty, cd /var/log, then ls to list the files there. It is a place to start.

Optionally, looks like you may need to reload the SD card and install one of the applications at a time. Looking at your log, maybe a day between installing things. The graphs1090 is the tool we all use and glad it helped you see what has been happening. We will be here to help as we can. Have fun with this.
Gene

@foxhunter and @astrodeveloper This Pi is dedicated to the task.

Will use a cron job to reboot for now, or should it just be the process?

Will check the logs.

edit: removed own stupidity, sorry fox :slight_smile:

htop might help … shift-M to sort by memory use.
df -h will tell you usage also for tmpfs mounts (/run or whatever ram logging you’re using).
if it’s /var/log or /run you can use ncdu on those folders.

2 Likes

What happened? I can’t remember :wink:

1 Like

a cron job does not solve the problem, it only hides it.
I would midterm check what causes the memory usage.

100% agree.
But limited what I can do. It’s dedicated to this task, fully up to date.
From a quick 10mins looking it seem to be VRS consuming the resources with more and more child process being left open.
Seems each tab is one, on each client.
But it was just a Quick Look.

10 Mins are not enough to monitor memory usage which increases over days.
VRS is pretty ressource intensive, but it should stop at some point.

You could shut down VRS for a day and see what happens.

HI @foxhunter.
10mins was probably a poor way to describe it… 10mins to investigate and try a few things:
Normal usage, then a drop when i stopped the feeder
Rise when its restarted,
Then a large drop when restarting VRS, its been stable since.

image

Will now monitor…

2 Likes

Yeah, as said, VRS doesn’t care about memory that much. And it requires the underlying Mono framework which increases the usage.

What is this viewer? Sorry for irrelevant question in middle of other discussion.

It’s Virtual Radar Server (VRS). There is some discussion about it here:

1 Like

Thank you so much. I have set up and configured few minutes back in my computer… seems such a nice tool tool. Thank you for sharing the information…

1 Like

It is quite a fascinating bit of software, can give more details than FA or FR24 if configured correctly. Plus good for sharing feeds.

Geoff