Tool/software: TI C/C++ Compiler
Team,
Our devices were working fine on sp_3.13.0.3_2.0.0.0_2.2.0.6, until recently we found 2 devices (out of 100+ devices in production) getting disconnected from MQTT server at regular intervals. We have suspected the issue to be PS Poll as it happens on few routers and to disable PS Poll, the SP has been updated to sp_3.17.0.4_2.0.0.0_2.2.0.7. Disabling PS Poll did make any difference (querying PS Poll status was in-fact failing) and then we decided to update our WLAN power policy to always ON. This change also did not resolve our disconnection issue, however since we were using line powered devices, we thought the always ON policy would be a better option. Hence we rolled out both changes to our customers (SP upgrade and setting power policy to ALWAYS ON). However with these changes we are noticing few devices having issues with server connection i.e MQTT never connects successfully. We do have watchdog set to reset our device after 6 hours of MQTT connection failure and this also did not work somehow. The other threads were working fine and the device is definitely functional except the MQTT connection. Ultimately we had to hard reset few devices and they got connected successfully. In our logs we also saw few other devices which are offline for few days which is a serious concern as we cannot access them directly.
We are not sure if this issue is caused by the SP upgrade or the Power Policy change. As a quick work around, is it safe to roll back these changes i.e downgrade SP to sp_3.13.0.3_2.0.0.0_2.2.0.6 and set the power policy to NORMAL? Prior to the update, our devices used to recover from any issues with the help of watchdog. Now it seems like either the watchdog has stopped / failed or the NWP is not completely reset or recovered from a crash, by the watchdog reset.
Appreciate if someone can quickly look into this and help us.
Regards,
Zac