Asset Tracker V1 Spontaneous reboots

[billywalton]

Hi,

I am running the asset tracker v1 app with a few mods, but it seems to reboot once in a while and I cannot find the cause.

The LOG gives nothing away:

[01:07:55.576,293] [0m<inf> asset_tracker: RRC mode: Idle[0m
[01:07:56.069,366] [0m<inf> asset_tracker: RRC mode: Connected[0m
[13] -3,0,0,-1,-5,-4,-5,0,1180314815
[14] -7,0,-1,-2,0,-3,0,-2,1180315825
[01:08:03.826,904] [0m<inf> asset_tracker: RRC mode: Idle[0m
[5] -2,-2,-3,-3,-5,1,-1,-2,1180322899
[01:08:04.661,743] [0m<inf> asset_tracker: RRC mode: Connected[0m
[6] -1,-1,-3,-3,0,-1,-2,-2,1180323910
*** Booting Zephyr OS build v2.4.99-ncs1  ***
[00:00:00.202,331] [0m<inf> at_host: UART check failed: 12. Dropping buffer and retrying.[0m
[00:00:00.212,951] [0m<inf> asset_tracker: Asset tracker started[0m
[00:00:00.213,409] [0m<inf> watchdog: Watchdog timeout installed. Timeout: 60000[0m
[00:00:00.213,867] [0m<inf> watchdog: Watchdog started[0m
[00:00:00.214,233] [0m<inf> watchdog: Watchdog feed enabled. Timeout: 30000[0m
[00:00:00.224,487] [0m<dbg> nrf_cloud_transport.nct_settings_set: Settings key: p_sesh, size: 4[0m

Is there a way to track a reboot cause?

Thanks

[Albrecht Markus Schellenberger]

Hello Billy,

A good starting point could be to check the reset reason register.

You can do that via the debugger interface using the following bash-command:

$ nrfjprog --memrd 0x40005400

I hope this will help you!

Regards,

Markus

[billywalton]

Hi Markus,

Thanks, I am running this under the debugger, which is whats strange, as if there was a fault I would have expected it to stop execution

But It just continues as nothing is wrong....

Is this normal behaviour?

[billywalton]

Hi Markus,

After it restarts, i usually see the following in the log

[00:01:05.799,316] [0m<inf> asset_tracker: LTE cell changed: Cell ID: 131288194, Tracking area: 1392[0m
[00:01:05.906,982] [0m<inf> asset_tracker: RRC mode: Connected[0m
[00:01:11.716,094] [0m<inf> asset_tracker: RRC mode: Idle[0m
[00:01:11.847,137] [0m<inf> asset_tracker: RRC mode: Connected[0m
[00:01:16.878,204] [0m<inf> asset_tracker: RRC mode: Idle[0m
[00:01:17.013,397] [0m<inf> asset_tracker: RRC mode: Connected[0m
[00:01:18.433,776] [0m<dbg> nrf_cloud_transport.nct_mqtt_connect: mqtt_connect failed -104[0m
[00:01:18.434,295] [1;31m<err> nrf_cloud: nRF cloud connect failed -104[0m
[00:01:18.434,722] [0m<dbg> nrf_cloud.api_event_handler: NRF_CLOUD_EVT_TRANSPORT_CONNECTING[0m
[00:01:18.435,180] [1;31m<err> nrf_cloud: nRF cloud connect failed -5[0m
[00:01:18.435,607] [0m<inf> asset_tracker: CLOUD_EVT_CONNECTING[0m
[00:01:18.435,974] [1;31m<err> asset_tracker: Failed to connect to cloud, error -5[0m
[00:01:18.436,431] [1;31m<err> asset_tracker: Device will reboot in 300 seconds[0m
[00:01:18.436,889] [1;31m<err> asset_tracker: LTE link disconnect[0m
[00:01:19.912,231] [0m<inf> asset_tracker: RRC mode: Idle[0m
[00:01:20.049,072] [1;31m<err> asset_tracker: Shutdown modem[0m

and the cloud will not reconnect until I cycle power

[Albrecht Markus Schellenberger]

Thanks for the feedback, Billy!

Yes, in general the application will restart after a reset. Normally you can prevent this by adding:

CONFIG_RESET_ON_FATAL_ERROR=n

to your prj.conf.

The error handling of the asset tracker is a little bit different, though.

Do you face the same issue when you are running the application without using the debugger?

Regards,

Markus