OK, very good. May be worth adding to your instructions, or at least make it more clear that Step 6 is only required when using both dump1090-fa & dump978-fa. You do say this, but it may not be clear to dummies like me.
OK, I see that the original value of the device index was 0, not 1000. I changed it back to 0. Seems to be working.
Again, many thanks!!!
You made a mistake to follow above instructions, which are under following heading.
6 - Configuration for using both dump1090-fa & dump978-fa
It does not apply to your install as you did NOT install dump978-fa.
No need to add any further instructions, as it is very clear from the heading of the item when to use it.
2 Likes
Agreed. OK, many thanks again!!
I do notice somewhat better results going from piaware v3.6.3 on Debian 9 to piaware v3.8.1 on Debian 10.
Congratulations that you succeeded in upgrading your system (OS+piaware+dump1090-fa) to their latest versions. 
1 Like
I could have NEVER done it without your EXCELLENT instructions!!! Every time I go through the exercise of building piaware from source, itās a great learning experience. I guess Iāll be doing this about once a year, which means Iāll be bugging you for the latest build instructions on an annual basis!!
2 Likes
Well @obj possibly can shed some light on this.
1 Like
I still have the old VM, so I can compare the two. But I will be sticking w/ the new version.
There have been two major changes. May be improvement is combination of both.
- Your OS has been upgraded to
2020-02-12-rpd-x86-buster.iso(which is based on Debian Buster 10.3) - Piaware+dump1090-fa upgraded to v3.8.1
1 Like
top - 08:24:27 up 1:47, 3 users, load average: 0.44, 0.50, 0.48
Tasks: 140 total, 1 running, 139 sleeping, 0 stopped, 0 zombie
%Cpu(s): 18.4 us, 16.7 sy, 0.0 ni, 64.3 id, 0.0 wa, 0.0 hi, 0.7 si, 0.0 st
MiB Mem : 431.3 total, 12.6 free, 139.2 used, 279.4 buff/cache
MiB Swap: 610.0 total, 610.0 free, 0.0 used. 275.5 avail Mem
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
20379 dump1090 15 -5 29348 7656 3120 S 23.5 1.7 9:13.43 dump1090-fa
23349 piaware 20 0 17072 9392 5864 S 1.3 2.1 0:18.43 fa-mlat-client
9 root 20 0 0 0 0 S 1.0 0.0 0:20.08 ksoftirqd/0
606 www-data 20 0 10112 5420 4456 S 1.0 1.2 0:36.20 lighttpd
23293 piaware 20 0 27592 12312 6976 S 1.0 2.8 0:15.07 piaware
31581 pi 20 0 10644 3736 3220 R 1.0 0.8 0:00.21 top
10 root 20 0 0 0 0 I 0.7 0.0 0:25.98 rcu_sched
1143 pi 20 0 144508 30116 24780 S 0.3 6.8 0:09.58 lxpanel
2551 pi 20 0 12440 4660 3816 S 0.3 1.1 0:02.06 sshd
23312 root 20 0 0 0 0 I 0.3 0.0 0:00.26 kworker/u2:1-events_unbound
23346 piaware 20 0 3948 2260 2072 S 0.3 0.5 0:04.80 faup1090
1 root 20 0 17008 8880 7148 S 0.0 2.0 0:14.40 systemd
2 root 20 0 0 0 0 S 0.0 0.0 0:00.03 kthreadd
1 Like
Question: Does sending feeder output to addition sites (e.g. adsbexchange, using their āscript methodā) impose significant performance overhead?
I have it running on a Raspberry 3B and feed overall six sites. I cannot identify any performance issues.
1 Like
There were 160 commits between 3.6.3 and 3.8.1: Comparing v3.6.3...v3.8.1 Ā· flightaware/dump1090 Ā· GitHub
So yes Iād expect them to behave differently.
3 Likes
Thanks for the info, @obj. 160 commits in approx 16 months, so about 10 commits per month. Out of those, I can believe thereās at least one or two major enhancements ;-).
Good to know, @foxhunter . Given that I have mine running on a VM on an 11 year old Netbook with an Intel Atom N450 1.66GHz single core CPU (remember those?), I think Iāll keep the overhead to a minimum. So I wonāt be adding any additional feeders.
Most of the feeder are taking the data delivered by your dump1090 instance. So the CPU load is not increasing that much.
Just checked it. While dump1090 is currently using 15-20% CPU, the other feeders and clients are at 1-3% each.
1 Like
@foxhunter Thanks! Appreciate the info. Like I said, my setup is running on a very wimpy platform. I find it endlessly fascinating that it works at all. I will probably keep mine as is. But if I ever upgrade to an RPi, I will probably add some additional feeder(s).
There are people running it on a Pi Zero which is a single core CPU with 512 MB RAM
2 Likes
OK! Good to know. Thanks.
OK, I finally added adsbexchange to my feeder using the script method on my FA feeder. As you said, there is negligible overhead. Itās working fine!! Running it in a Linux virtual machine w/ 512MB memory on an Asus Netbook Eee PC 1005-PEB ānetbookā (circa 2009) with an Intel Atom N450 CPU and 2GB memory that is running Windows 10 āStarter Editionā (32-bit).
What i have identified is that ADSBx causes more traffic via Internet than FR24 or FA feeding.
Just in case you have a metered connection. If not, you do not need to care.about
1 Like