The Serial diagnostic stream is raw intelligence - needs context and interpretation. Those timeouts were first try errors and must have been recovered. The Serial datalog isn’t timestamped, so in this scenario doesn’t tell me as much as when I’m watching it real-time.
Pull out your AC reference plug and you’ll find out.
It’s a diagnostic message I forgot to pull out. It means that when explicitly sampling a voltage channel, as opposed to doing it as part of a power channel sample, the result returned was zero. That can be caused by the AC reference being pulled out, or the sample routine could be returning zero because it detected that there was an interrupt during sampling which would invalidate that sample.
If you look at the period from 17:04 to 1:00 the next day, there were 11 of them. AT minimum 2 explicit voltage samples per second (in a fully loaded system), that means 11 out of about 58,000 explicit voltage samples might have been interrupted. That’s .02% and perfectly normal.
. Even custom port numbers withing the URL are honoured.