I generally tabulate my daily measurements just after midnight, and open up the sense app on an iPhone to do so. I have noticed, in general, that in the first five minutes after midnight (Pacific Time), the app often responds with timeout/server errors (“Request failed. Your network could be down, or the Sense server returned an error. Tap to try again”). It usually doesn’t last long.
Tonight, however, its been over 15 minutes. The status page says everything is OK. Making my own API calls returns this:
My network is, obviously, not down as I do receive a response from the server (a 502 error with the bad gateway message). Something in the server configuration, however is not alright.
I also tried accessing the web app, and it fails similarly, but doesn’t give any error messages. The orange bubbles just keep on bubbling forever…
The usefulness of the error message is suspect. It should not be hard (at the coding level) to detect whether there is no response from the server (which might indicate a network problem), or an error response (http status not 200). Please present a different error message for these scenarios.
Furthermore the usefulness of the status page is also suspect. It is, of course, possible that only I am experiencing this problem, but the nature of the problem strongly suggests otherwise. To me that means the status page’s underlying monitoring system is inadequate. It may not detect the same difference between no connectivity at all vs. error responses, or it may not even try that (perhaps just doing pings), or there is some built in delay before a failure is causing it to be marked as broken…
Eitherway this calls for improvement…