Summarize of the feature request

Current behavoir:

When a controller fails to publish a message the connectionFailures variable is incremented. And if the connectionFailure count exceeds the Connection Failure Threshold setting the system reboots. Unfortunately the connectionFailures value is a persistant count that requires a reboot to reset back to zero.

This is an issue since random connections failures can occur in a stable system due to a busy MQTT controller, temporary WiFi signal issues, etc. Over a period of days or weeks the count will eventually exceed Connection Failure Threshold setting. And a unwanted reboot will occur.

Of course the Connection Failure Threshold setting_ can be disabled and not used at all. But in some situations it would be desirable to force a reboot after too many sequential failed publish attempts have occurred.

Expected behavior

The connectionFailures count should be reset to zero after the system has successfully reconnected. The reset should be qualified by requiring two or three successful connections in a row before clearing the count. This new behavior could be implemented as a checkbox option on the Advanced settings page. Perhaps something like "Non-persistent Failure Threshold [ ]"

• Thomas