All devices/sensors read null for a single interval

It seems every device/sensor/etc. in the database returns NULL for a single polling interval, and then it returns to normal. So while it works for 287 out of 288 checks a day, that one errant check sends our engineering team and CEO about 1000 scary emails kicking all the response teams into action. I updated to 21.10 two weeks ago, shortly after it was released and there has been no issue until yesterday and today. I looked in the cron listing and didn't see anything running around that time.

Love this tool, thanks so much for all of your hard work. If there is anything else I can pull to help close this issue out please let me know.