Admittedly this is not a project, it’s a recurring problem I’m having, but after spending the better part of half an hour to figure out where to post it, this is the only section vaguely close.
So here we go: at random my HA interface gets inaccessible both from the browser as well as from my iOS app.
The only symptom I was able to find is the manifest.json
stuck on pending.
If I ssh in my pi to check if any process overloads it, but nothing:
-
htop
shows all cores chilling in the low single digits percent. -
docker logs -f homeassistant
, as well as the other containers logs, are silent or at most showing access logs i.e.:
21-12-28 16:47:36 INFO (MainThread) [supervisor.api.proxy] Home Assistant WebSocket API request initialize
21-12-28 16:47:36 INFO (MainThread) [supervisor.api.proxy] Home Assistant WebSocket API request initialize
21-12-28 16:47:36 INFO (MainThread) [supervisor.api.proxy] WebSocket access from a0d7b954_nodered
21-12-28 16:47:36 INFO (MainThread) [supervisor.api.proxy] Home Assistant WebSocket API request running
21-12-28 16:47:36 INFO (MainThread) [supervisor.api.proxy] WebSocket access from a0d7b954_nodered
21-12-28 16:47:36 INFO (MainThread) [supervisor.api.proxy] Home Assistant WebSocket API request running
21-12-28 17:06:58 INFO (MainThread) [supervisor.homeassistant.api] Updated Home Assistant API token
This seems to happen at random and the only “fix” I found so far is to keep rebooting the pi until HA starts working again. Not very helpful to understand what/why.
I haven’t added a new device to my home for ages, and the whole system might work fine for weeks, and then all the sudden boom, this happens.
Anyone perhaps experiencing the same issue and/or having any idea on what might cause this?
I also moved my recorder to maria DB, with a DB installed on a totally different machine to exclude a possible painful point but it didn’t help at all.