-
Notifications
You must be signed in to change notification settings - Fork 676
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[Bug]: Factory reset on 1.3 causes problems #1597
Comments
Problem is, without a repo or a serial log this is near to impossible to debug. I have tried your steps and can not reproduce it on first try. |
I'll see what I can do. |
I've seen this on the TLORA board as well. Reflashing it solved the issue, as you indicate. |
@caveman99 I get weird behavior on every factory reset with every device, on NRF it does not seem to clear any settings with 1.3.30, on a tbeam it makes the device really unstable and tlora devices are unstable but don't boot loop. |
RAK 4631 ??:??:?? 6 Installing default DeviceState ??:??:?? 14 Expanding short PSK #1 The device crashes when it gets to Loss of power in Powered, device has a charged battery and is connected via usb |
On the road for the weekend. Does that also happen with canned message module disabled?
|
Here it is with no canned messages 4631 Rak with buzzer, GPS and BME680 WARNING file:mesh_interface.py _sendToRadio line:467 Not sending packet because protocol use is disabled by noProto ??:??:?? 13 Expanding short PSK #1 |
I am also seeing a bootloop (stuck on meshstatic logo) after calling the factory reset command using python CLI. Before the reset was issued, the device had just been reflashed and minimal settings configured: owner, Lora region, channel config. meshtastic-python: meshtastic/python@03aab10 I'm assuming that without a GPIO serial debug setup enabled, I will be unable to produce the necessary logs. Is this correct? |
Also happens on my heltec 2.0 board |
I believe the bricking issue (at least with CLI) is that the node doesn't report max_channels after factory reset: Factory reset device
Freshly flashed device
The CLI uses max_channels reported by this protobuf to determine whether the device has an appropriate version. This could however just be a symptom as nothing on the device itself appears to have reset:
|
I have seen zero cases of bricked devices, all have been recoverable by refreshing the device. It is really hard to brick an ESP32 dev board. |
By bricked I mean it can't communicate with any of the apps, flashing still works. |
People get nervous when they hear bricked, generally means the hardware is done for. |
When trying to use factory reset to disable WiFi I get boot loops, it still thinks WiFi is enabled and has a region set as well |
We've had issues with run time of garbage collection on LittleFS before, maybe it's resetting before the device completes the write cycle so it still sees the old files on bootup? |
another approach to factory reset would be to simply loop the flash FS area and null it. Not use the FS functuions at all. LittleFS will init the flash area if it doesn't find a valid partition marker. |
Waiting longer before reboot might also work. When writing #1653 I initially had at 15 second reboot to get stuff to fully clear, but then decreased to 5 and it still seemed to be working. |
Is this still a problem with the current master build? ref. https://github.com/meshtastic/Meshtastic-device/suites/8445496242/artifacts/374421783 |
meshtastic --factory-reset has been working reliably for me on the latest builds. Closing for now |
I believe I'm still having related problems, but it's hard to tell because in my T-Echo's current state, it bootloops on every version of Meshtastic I try, after calling factory reset. Since I can't get it to not bootloop I can't reset it. It still runs other firmware life SoftRF fine, but even 1.2 can't cope with the state the config is in now. |
There is a reset .uf2 file that is pinned to the #nrf52 channel on discord, try dropping that on to the drive and then connecting to it via serial (meshtastic --noproto) That will wipe all saved data from the device. |
Thanks! That did allow me to recover from this bad state. |
I think we need to either build a fix into the firmware or make the step of clearing the flash more obvious |
since using meshtastic --noproto I am stuck in a loop where the serial port sends out the "press any key to contiune factory reset" it says "Formating ..... Done!" but then on reset or rest via button it loops back to "Press any key to continue factory reset" |
Category
Other
Hardware
T-Beam, T-Echo
Firmware Version
1.3.30.9fe2ddb
Description
I've observed two different problems after asking 1.3.30 to factory reset (
meshtastic --set device.factory_reset 1
).On T-Echo, a factory reset will cause it to bootloop. Downgrading to 1.2.25, setting the region, and then upgrading to 1.3.30 again works around the issue.
On T-Beam, a factory reset causes the device to be unable to communicate via serial or bluetooth. In both cases the error received is that the firmware is the wrong version and needs to be upgraded. Re-installing the firmware with
device-install.sh
works around the issue.These are the only devices I have, so I can't verify what happens elsewhere.
I'm unable to get any serial logging from either device when in the corrupted state.
Relevant log output
No response
The text was updated successfully, but these errors were encountered: