I’ve just done an upgrade to Home Assistant OS 6.6 and now all my ESPHome devices are not connecting.
All of them show like this:
Is anyone else seeing this, or am I alone? Any idea how to fix this problem, please?
I’ve just done an upgrade to Home Assistant OS 6.6 and now all my ESPHome devices are not connecting.
All of them show like this:
Is anyone else seeing this, or am I alone? Any idea how to fix this problem, please?
I upgraded to OS 6.6 about a week ago without any problems. Some 20 ESPHome entities here that all survived. What’s in your ESPHome node log?
Thanks, @Jpsy - it appears that ESPHome is communicating with the nodes as I can, for example, see their logs from ESPHome. The problem seems to be that ESPHome and HASS are no longer communicating because the devices ESPHome sees are no longer available in the HASS front end and it sees them as orphans.
OK, mystery solved - sort of. The upgrade of OS has caused HASS to create new instances of all the ESPHome devices and and increment the numbers (e.g. cover.kitchen_blind is now cover.kitchen_blind_2).
Very weird and frustrating but relatively easy to correct, thank god (I have nearly 100 ESPHome nodes!)
Check your entities in HA. I bet for some reason HA has created new entities for your ESPHome sensors with “_2” appended to the entity_id.
If so:
Delete the old orphans and rename the entity_ids of the new copies to remove the _2.
Thanks - I think our posts crossed!
Yepp!
You could also stop the core and edit the entity storage file directly.
I would if I knew how! That sounds like something a relatively inexperienced person like me could get VERY wrong!
True.
It will be safer to delete the old and new instances of all affected integrations from HA and restart HA. That should re-discover them with their original entity names (untested!).
The really weird thing is that the _2 change is totally inconsistent. Some entities have _2, some are unchanged - on the same device!
Very strange!
The affected nodes might have changed their MDNS names on the network (i.e. “node-name.local”). I vaguely remember some change in handling underscores in node names. You should avoid underscores.
Update:
Nonsense. Ignore. That would always affect ALL entities of a node.
I use reserved IP addresses on the DHCP server for all nodes and no changes to their names. All very strange but thanks so much for your help!
And I’ve now discovered it did not affect all nodes! It could hardly be more inconsistent!
I am also experiencing issues with all my ESPHome devices. I am not seeing any renamed entities and some of the devices works and some not. In the ESPHome add on, all the devices are shown as offline.
I have some TP-Link / Kasa devices and Google Nest camera that also no longer works after the upgrade.
Hope we can sort this out soon!..