search

newracom / nrc7292_sw_pkg

NRC7292 Software Package for Host mode (Linux OS)
http://www.newracom.com/product
GNU General Public License v2.0
58 stars 27 forks source link

overrun, tx size:151508 #63

Open synaption opened 1 year ago

synaption commented 1 year ago

Under heavy or moderate use, the mesh point portal is showing the message overrun, tx size:151508 many times. Sometimes this causes the network to stall. Sometimes it is ok. It seems like some devices may be worse than other which could point to a hardware issue. I am wondering if there is a way to solve this. Perhaps by increasing the size of whatever is overrunning.

synaption commented 1 year ago

I have a 10 node mesh up and under some circumstances I can also get one of my mesh points displaying this message which does corelate with the rest of the network to stalling. I'm wondering if having them all in the same room is part of the problem. Any advice on how to keep one bad apple from spoiling the rest?

newracom commented 1 year ago

Dear @synaption,

Thank you for your questions. The following feature was added on v1.3.4 rev13, https://github.com/newracom/nrc7292_sw_pkg/releases/tag/v1.3.4_rev13

Could you try with the latest package and share the result?

Best regards, Newracom

synaption commented 1 year ago

So I got around to looking at this again. Suppressing the log does not really solve the problem. This log info appears when the network is not functioning. Currently None of the nodes will connect. The CPU seems to be set at my devices max frequency by default. I think all the devices are spamming up the network trying to connect at the same time. I could have them start more gracefully one by one, but I would have to script that since I don't have ssh access without the network functioning properly. 

$ cat//sys/devices/system/cpu/cpu0/cpufreq/scaling_available_frequencies 
1000000 1500000 
$ cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq 
1500000
$ cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_max_freq 
1500000