Problem
I'm experiencing long cycle times, ~60 Seconds instead of 5 Seconds, because of a powered-off Modbus-Slave.
Situation Specifics
I have 3 Modbus RTU and 3 Modbus TCP slaves connected to my thin-edge device. when they are all work properly I poll each device every 5 seconds. Sometimes one or more of the RTU Slaves get powered-off. In these cases the time used to poll the Slave rises to 60 Seconds per powered off device.
modbus.log
Culprit
Based on the Logs we can see that every poll request takes 3 seconds before throwing an exception.
Ideas
- reduce the timeout or make it configurable
- stop polling the device if it doesn't answer