Traces and remote admin does not work after few minutes turn on/reboot

[xejarque]

I'm using the latest version 1.17.3 on my heltec v4.3 repeater.
The problem:

• Turn on or reboot the device and traces and remote admin for repeaters more than 1 hop are working as expected.
• After few minutes, strong traces with just 1 hop not working anymore (timeout) and remote admin neither can access.
• Reboot and work few minutes until fail again
  Very anoying... Any idea?

[M4rc-Andre]

I'm having the same problem with a Heltec V3 at a remote location that I can no longer manage. I can still log in. I can access telemetry and neighbor data via the app. But none of the CLI commands work anymore. Other than that, it's still repeating, but I can no longer configure it. I also can’t get to the device quickly enough to flash an update via USB. It’s very frustrating because “start ota” doesn’t work anymore either.

[liquidraver]

zephcore was created for european laws in mind, so duty cycle is 10% by default.
I think you guys hit the duty cycle, admin packets are deliberately left out from the token bucket.

can you try to give it "set dutycycle 50" and test it after?

[liquidraver]

also thinking about zephyr's power management, I had a lot of trouble with it, I'll check

[xejarque]

I think that my problem is different to @M4rc-Andre commented, I have my repeater with zephcore in the roof, I can connect to it and change whatever without problems.
I still can do ping and trace to the repeater, the ping have a very long delay, sometimes until 5 seconds and is not stable as original firmware that is instant.
The problem is:

• when I use my repeater to perform a trace from my campanion to another node. It does not work.
• when I try to admin other remote repeater using my repeater in the path. example a repeater 1h away and I have to use my own repeater to reach them.
  Another extrange thing is when I send a flood message, sometimes I'm not getting my repeater heard the message, and is +12db from my companion.
  If I reboot the node work properly until some minuts 🤷‍♂️

[liquidraver]

do you guys see "Debug flags 2" in a few minutes on the stats?
I have a theory, power management kicks in 10 minutes after boot, so I think it starts there. if debug flags immediately set to 2 then it's an indicator I'm heading down the right lane in the code

[liquidraver]

also could you check if reported uptime vs real elapsed time is the same?

[M4rc-Andre]

Sorry about that. I drove the 50 km there and back and reflashed the affected repeater with the original Meshcore.io firmware. It’s just not good when things stop working that far away. It’s frustrating because I was just there two days ago and installed the update. Fuel isn’t cheap in Germany right now.

[liquidraver]

yeah, I never play with zephcore where I'm not able to update it frequently or reach it by serial somehow

[xejarque]

Sorry for late response, I also decided go back to the original firmware because I cannot live without remote management 😢
I use zephcore with my t1000 companion and it is working perfect, but with repeaters I cannot take the risk of lost remote admin neither traces.