Or is it just mine? Seems like pretty regularly, the sensor reboots every 20 minutes (timings derived from uptime reported in /json endpoint) - sometimes it’s off by a few minutes, but
10:02:19
09:42:33
09:25:46
09:05:00
Thoughts?
Huh, odd … after monitoring things for a few days … it suddenly stopped rebooting itself yesterday. The graph below shows two things -
- before ~2pm on Dec 27, the sensor was just in a fast reboot loop (see other topic) every 30s.
- from ~2pm on Dec 27 until ~noon on Dec 29, the sensor was rebooting every ~20min …
after ~noon Dec 29, there’s been no indication of reboot and to support this, the uptime field in the /json payload shows that it’s been up for 28h
When a PurpleAir sensor can’t communicate with PurpleAir, it will reboot once every 20 minutes. From Dec 27 to Dec 29, there was little or no communication from your sensor to our systems, causing these periodic reboots.
1 Like
Gotcha, thx @Josh_PurpleAir. Is there output in the serial logs when that’s happening? I don’t recall seeing anything super obvious - I saw the JSON payload it was posting (trying?) to www.purpleair.com:443, or is that something else?
There’s currently no indication over the serial port indicating the reason for the reboot, though this is something that could be added.
The JSON payload you saw is the 2-minute report the sensor sends to us.
So it wouldn’t log any error code or similar if the 2-minute report didn’t send successfully?
If the 2-minute report doesn’t successfully send, there should be two lines of serial output:
- “Failed” with a number representing the milliseconds between the send attempt, and when a failure to send was determined
- “Error” with a negative number indicating the error
This is currently only reflected in serial output, and not in the locally served JSON data. Other fields in JSON do exist which can indicate failures to send, such as httpsuccess, httpsends, and the status indicators.
Cool, thanks - if I see it happen again, I’ll have something to look for. Appreciate the response!