Sparrow firmware hangs intermittently

[zfields]

I have noticed on several occasions that my Sparrow device will stop checking in. When the Sparrow gets in this state it appears to be completely seized up.

When it hangs:

• It will not respond to the PAIR button.
• It will not respond to t+Enter on LPUART.
• It will not respond to the debugger.

Interestingly, after a non-deterministic period of time, the Sparrow will resume it's standard behavior (without restarting).

I have logged the memory usage, and there does not appear to be a leak.

Unfortunately, I cannot deterministically reproduce this issue at this time.

[m-mcgowan]

Should we use the watchdog to help guard against hangs where the main processing loop is blocked for an extended period? (e.g. a few minutes.)

Naturally, we want to find the cause of the issue, although using the watchdog will provide a fail-safe fallback until this hang is fixed and for any other unknown stall bugs that we may have.

[zfields]

Yes. I agree a watchdog should be implemented regardless of the outcome of this issue. This consequences of this behavior are too dire with regard to battery life and the overall effectiveness of the system.

As always, the only nagging concern is... Would we have ever noticed this behavior if the watchdog was already in place?

[zfields]

Perhaps the "satisfactory" answer to my question is to improve logging, so we have more visibility into, and/or opportunities to catch, misbehavior.