Allo people
I’ve had HA happily running for a couple of months now, I’ve been loving it.
However, I woke up the other day and it was unnacessible.
I’m running Hass.io in a Linux VM.
It looks like the HA Docker container isn’t connected to a network, but I’m not sure if it should be, considering the supervisor is connected?
I don’t think home-assistant.log
shows anything useful.
2018-12-20 18:27:23 WARNING (Recorder) [homeassistant.components.recorder] Ended unfinished session (id=23 from 2018-12-20 14:32:53.178395)
2018-12-20 18:27:34 WARNING (MainThread) [homeassistant.components.light] Setup of platform flux_led is taking over 10 seconds.
2018-12-20 18:27:43 WARNING (SyncWorker_4) [homeassistant.components.light.flux_led] Failed to connect to bulb 10.50.11.12, raf_strip01
2018-12-20 18:28:05 ERROR (SyncWorker_3) [pychromecast.socket_client] Failed to connect, retrying in 5.0s
2018-12-20 18:36:01 ERROR (Thread-7) [pychromecast.socket_client] Error reading from socket.
2018-12-20 18:36:01 WARNING (Thread-7) [pychromecast.socket_client] Error communicating with socket, resetting connection
2018-12-20 18:36:01 ERROR (Thread-7) [pychromecast.socket_client] Failed to connect, retrying in 5.0s
I’m aware of the third line, I have that turned off while HA isn’t working, and I can’t imagine the pychromecast failure would kneecap HA like this.
Is any able to lend a hand?
I’m a Docker noob but can make my way round a Linux system without much trouble.
What makes you think this? By default it is connected to the host for network mode. Unless you changed it, it’s still that way.
I’m unsure on Docker networking, but from the IPs on other containers I assumed it ran an internal switch and NAT-ed things out to the host network, as 172.x.x.x isn’t part of my IP schema.
Things that are using --net=host
will not show an IP or published ports. This is normal.
Ah OK, cheers dude. Learning more about Docker minute by minute! Hadn’t touched it before Hass.io .
Do you have any idea where else I could look for this issue?
Have you thought about checking your router to make sure it is still forwarded to the right IP?
Yeah, the DNS record is up to date. I can still access the configurator addon at port 3218 as well.
Bump. Unless anyone is able to lend a hand here, I’m thinking I’ll just blow away the container and reinstall, then restore my config.
You can still access HA locally right?
If that’s the case, HA isn’t the issue here and you need to look at your network configuration.
No, I can’t access HA at all.
Even by using your local IP?
Yep, not even by using the local IP.
There has to be something in the logs. Open a command prompt and run docker logs -f --tail 0 homeassistant
when you restart the container
I can’t see anything that indicates an issue in the logs, but I’ve pasted them here in case you can see something I’ve missed.
https://pastebin.com/X4Q8gGMH
Looks like it’s actually dying at the Chromecast connection. Do you have Chromecast set up in config? Try removing it? Try removing discovery?
leinich
December 23, 2018, 4:08pm
16
Could be you are currenly facing the same issue I had yesterday.
Once I disabled chromecast the frontend showed up again.
flamingm0e:
Try removing discovery?
Commented that out and it appears to be doing the same thing.
How can I disable pychromecast? As above, I’ve disabled discovery.
leinich
December 23, 2018, 4:52pm
18
This was not working for me either
I configured Chromecast via UI, so I need to undo this.
As there was no UI, I used ssh and removed it from the configuration.
There is a hidden folder under config
/config/.storage/
There I removed the chromecast part in core.config_entries
nano /config/.storage/core.config_entries
Holy shit that fixed it, cheers dude.
Do you know if there’s a fix for this, other than disabling pychromecast? It’s a pretty important part of HA for me.
Check this out. Lots of people saying it fixed the issue. Currently looking for that directory.
https://github.com/home-assistant/home-assistant/issues/14956#issuecomment-397802817