Expose detailed status to HomeAssistant

[CoMPaTech]

Is your feature request related to a problem? Please describe.
Unless I overlooked it in my dev window (I've got a generous lot of entities), I didn't spot a status for room_assistant itself (I do see my bluetooth devices). As sometimes the PI0w's are unavailable (wifi intermission) or out of memory, or maybe just unplugged one to vacuum (heaven forbids), I'm not sure if tracking is accurate. While I could set-up monitoring in various ways, exposing a list of active nodes will surely help spotting issues easily.

Describe the solution you'd like
Add a room_assistant device with attributes describing each cluster node. Not yet looked at your code but adding a sensor.room_assistant or something similar with some attributes like:

device |  state | attributes
sensor.room_assistant | [on/off] | last change # last mqtt message?
 | | nodes: ["livingroom","kitchen"] # i.e. the nodenames in a list, actively updated
 | | leader: ["kitchen"]

Describe alternatives you've considered
Adding 'tracked by node' to each tracked device, but that will be cluttering (though useful when debugging, but you can just subscribe to mqtt for that)

Adding nodes as separate sensors, but not sure (besides showing things) if it would be useful in the long run (i.e. sensor.ra_kitchen with attributes like devices_tracked: 3, last_updated

Additional context

[mKeRix]

Good idea! Should be useful for understanding the tracking history and some issues. Building on your suggestion above and my understanding of how Home Assistant entities should be used nowadays I would think adding two sensors makes sense: sensor.room_assistant_cluster_size with the count as state and the nodes in the attributes and sensor.room_assistant_cluster_leader with the current leader as state.

[CoMPaTech]

That indeed sounds great!

[Alfiegerner]

Great idea, this would definitely be very useful disgusting issues - I had an situation where my quorum wasn't reached due to varied reasons and I couldn't figure it out.

[github-actions]

🎉 This issue has been resolved in version 2.5.0 🎉

The release is available on:

• npm package (@latest dist-tag)
• GitHub release

Your semantic-release bot 📦🚀

[mKeRix]

Quick side note: the sensors I suggested above are now registered per instance, i.e. you will have two new sensors for each instance in your cluster. Decided to go that route because it's much easier to find split cluster issues this way.