I’ve seen some references to this in much earlier topics, but no serious resolution.
I am using the new entity ids setup for Zwave, and have a zwave+ fibaro double switch fgs223. Adding this switch to my network appears to highlight a small bug in how the zwave component processes entities with multiple instance values. Both switches have the same entity id and are differentiated by their instance value. This makes sense, but what does not are the derived switch names.
The switches appear with the entity ids:
switch.switch
switch.fibaro_system_fgs223_double_relay_switch
This naming is unintuitive, and renaming the zwave node has no effect on them. My other switches are given a name derived from the name I gave the zwave node.
I’m also seeing inconsistent behaviour where turning the switch on/off from HA is not working, or the state of the switch being reported is not correct. I’m wondering if these issues are related to the strange naming mentioned above.
I’ve had the issue with each of the last two zwave switches I added. And yes when the problem occurs you need to use the name switch.switch in automatons and scripts until you fix (it unless there are more than one named switch.switch). The most recent case on my system was 2 days ago running on version 0.64.0. This time renaming the entity registry to entity.registry.yaml.old the and re booting fixed the problem. I considered considered editing the entity registry file but I hadn’t seen the documentation so I did not try that.
