We have a client script running on a machine which isn't actively monitored, which just turns on and off some lights (so not safety critical). However, we'd like to know if there are any issues while the script is running -- such as the serial port closing, or a calculation somehow throwing an unrecoverable exception.
We catch errors and log them to the gateway (through a message handler), but this relies on someone going into the gateway logs every so often to look at them.
Is it an acceptable idea to have the message handler, in addition to writing to the logs, check if the indicated logging level is above a warning, and then act on that (such as sending an email)? Or is there a better way to handle this