AWS-FreeRTOS - TICC3220SF-LaunchXL random reset during publishing data


  1. I am using aws-freeRTOS demo example with TI-CC3220SF-LAUNCHXL.

  2. The main application task captures sensor data approx 130kb bytes and publishes chunk by chunk to the aws cloud.(each chunk carries ~15000 bytes).

  3. The logging task and watchdog task is running with low-priority to capture the debug traces & reset the board(if any task getting stuck) respectively.

  4. The problem is, while publishing the sensor data, in between most of the time Board is getting reset without any indication. (OTA demo version 1.0.0 - start of the application from below traces).

Please find the following debug traces for more information and help me out with this issue.

`185 20825 [iot_thread] [INFO ][DEMO][20825] Publishing data bytes 0 to 15041.
186 20825 [iot_thread] [INFO ][DEMO][20825] Bytes available on heap: 35816
187 20828 [iot_thread] [INFO ][MQTT][20828] (MQTT connection 20009088) MQTT PUBLISH operation queued.
188 20866 [iot_thread] [INFO ][DEMO][20865] Starting publish to topic $aws/rules/00000474M/data/vibx
189 20866 [iot_thread] [INFO ][DEMO][20866] Publishing data bytes 0 to 15043.
190 20866 [iot_thread] [INFO ][DEMO][20866] Bytes available on heap: 34752
191 20868 [iot_thread] [INFO ][MQTT][20868] (MQTT connection 20009088) MQTT PUBLISH operation queued.
0 1000 [Tmr Svc] [INFO ][DEMO][1000] OTA demo version 1.0.0

1 1000 [Tmr Svc] Simple Link task created
2 1144 [Tmr Svc] [INFO ][DEMO][1144] Device came up in Station mode

3 1151 [Tmr Svc] [prvPAL_GetPlatformImageState] xFileInfo.Flags = 0250
4 1152 [Tmr Svc] [prvPAL_GetPlatformImageState] eOTA_PAL_ImageState_Valid
5 1155 [Tmr Svc] Security alert threshold = 15
6 1155 [Tmr Svc] Current number of alerts = 0

Hi, Suresh. There seems to be a lot to grapple with here. Have you tried disabling all the auxiliary tasks to see if the problem persists? It might be helpful to start with the simplest functionality, and then add back a little more after you’re satisfied that the last iteration was stable. You could start by trying to send a constant value to the cloud at regular intervals with the other tasks disabled.