Hi @bpwwer, The status of the NS in the AC is always True even if I stop the Node Server.  I've only waited a short time, does it eventually toggle to False?

Some NSs toggle immediately (VUE, Notification for example)

Others have the same behavior as WeatherFlow (August, MQTT)

Why do they act differently?

Thank you.

Some node servers set up a status value that tracks the status of the connection between the node server and PG3.  So when the connect is stopped, PG3 should send a update to the ISY/IoP with updated status. 

But not all. WeatherFlow sets the status as on-line when it starts but doesn't send any updates when it is stopped.  It was written before PG3 was updated with the ability to automatically track the connection status.

Thanks for the clarification!

11 hours ago, bpwwer said:

Some node servers set up a status value that tracks the status of the connection between the node server and PG3.  So when the connect is stopped, PG3 should send a update to the ISY/IoP with updated status. 

But not all. WeatherFlow sets the status as on-line when it starts but doesn't send any updates when it is stopped.  It was written before PG3 was updated with the ability to automatically track the connection status.

@bpwwer

Is there a post/list somewhere, or a way for us to tell which Node Servers update this node?

6 hours ago, DennisC said:

@bpwwer

Is there a post/list somewhere, or a way for us to tell which Node Servers update this node?

No, it's up to the node server developer to determine how it is handled.  It varies for a couple of reasons:

1) PG2 forced the node server to manage the connection state, while PG3 handles that internally.  So PG2 node servers that were ported without making any changes here will continue to try and manage it themselves and are probably only partially effective at it (like WeatherFlow)

2) The ability to configure PG3 to report the connection status to a specific node/driver was added after a number of node servers had already been ported and not all authors have gone back and updated the node server to take advantage of this.

I'm certainly guilty of #2 as I ported a lot of node servers over early.  As I make updates, I try to remember to make that change as well.