#installation-archived
74689 messages · Page 85 of 75
Updated perfectly and then restarted HA but after a restart it never worked so reinstalled HA again but it's now stuck
Did you have a backup of your HA instance?
Core
Yes
Did you delete your config folder?
Because, no part of an upgrade should have touched that
pip install --upgrade doesn't touch it
Yes now I have deleted everything and reinstalled but after installing and typing hass at the terminal its stuck
Why the hell did you delete the config folder?
It's also not "stuck":
- First start may be compiling python packages
- With no
loggerthe default is something likewarn, so the console output is pretty damn minimal
It's been over half an hour
You've tried the web UI?
Yes it's not accessible
You've checked top and ps to see if it's compiling (look for cc1 amongst other things)
On a Pi it's quite possible for upgrades to take 40+ minutes
Also, this is probably a great time to switch to Docker
Why are there 32 and 64 bit HAOS variants for RPi3/4?
Because historically there were only 32 bit variants, and some add-ons need 64 bit
GPIO only worked on 32 bit
Or HAOS if its a Pi and get something proper for running anything else on
AFAIK HAOS no longer supports GPIO so only the 64-bit variant would be needed.
(Was about to say isn't that the .5 version then realised its already may. Holy shit)
Home Assistant no longer supports GPIO natively
However, there are likely other reasons why people may want to continue using 32 bit - just because you don't know what they are doesn't make them invalid
Hi I am now stuck at the preparing Home
Did you check #installation-archived message ?
how can I fix time on rpi4 running HAOS ?
{{ now() }}
{{ utcnow() }}
``` shows
2022-05-05 00:42:33.635955+05:00
2022-05-04 19:42:33.636040+00:00
right now
Hello. I installed proxmox and HA on it. In proxmox I see really high usage of RAM for VM with HA - this is clean install without any addons. But in HA application I see really lesser usage of RAM.
Proxmox: http://prntscr.com/WeCfl6pmpNb7
HomeAssistant: http://prntscr.com/sd1eTE4Fif4b
Any ideas? Should I be worried about this?
I need a little help understanding a DNS issue I have since yesterday.
I am running Version: core-2022.4.7 /Installation Type: Home Assistant OS
Since yesterday it seems like a dns component updates itself. When running dns version 2021.06.0, everything works fine.
But then after "a while" (a few hours), it seems like that component updates itself to 2022.04.1 and from that time HA can't resolve any internal DNS entries. It seems like it just skips the internal DNS server and takes next in line. If I run ha dns downgrade --version 2021.06.0, everything works fine again. I can't find any info on this, so any help would be appreciated.
Additional info: If I change HA to use DHCP, it works with version 2022.04.1. But not when I use the static IP and DNS I have used for years.
No other changes, upgrades or so in the network
HI, I have HA installed on docker with supervisor and other docker images, two days ago I've installed watchower container for auto update my containers, but now HA show this error on logs "ERROR (MainThread) [supervisor.resolution.evaluations.container] Found image in unhealthy image list 'containrrr/watchtower' on the host"
Anyway, watchtower container is healthy, when I look into docker, so does HA explicitly look for specific containers and block execution of upgrade/restart if it find one of them ?
Are you running supervised install method then?
yes
Then yes as per the docs running non add-on stuff in HA breaks supportability. It is unsupported with other containers running
That arent add ons
I'm running more than 20 containers with HA since 2021, but this is first time when I see this error
➜ ~ ping pool.ntp.org
PING pool.ntp.org (162.159.200.1): 56 data bytes
64 bytes from 162.159.200.1: seq=0 ttl=55 time=135.741 ms
64 bytes from 162.159.200.1: seq=1 ttl=55 time=135.493 ms
64 bytes from 162.159.200.1: seq=2 ttl=55 time=135.501 ms
64 bytes from 162.159.200.1: seq=3 ttl=55 time=135.059 ms
64 bytes from 162.159.200.1: seq=4 ttl=55 time=135.421 ms
64 bytes from 162.159.200.1: seq=5 ttl=55 time=135.612 ms
^X64 bytes from 162.159.200.1: seq=6 ttl=55 time=135.174 ms
^C
--- pool.ntp.org ping statistics ---
7 packets transmitted, 7 packets received, 0% packet loss
round-trip min/avg/max = 135.059/135.428/135.741 ms
I can easily ping ntp server
https://github.com/hassio-addons/addon-chrony did the trick
@humble mirage I shouldn't tag, but I found the solution to time problem at least. 1. install chrony 2. enable "set_system_clock"🙂
@midnight steeple , it is specifically watchtower causing your problems
If you shouldn’t tag then… why tag
can anyone tell me how/where I can change the radius of the initial zone, that gets created during the installation? I found that the latitude data is in the DB file, and I can't edit the inital zone via the UI. On top of that I created a new test zone, and expected that it will be created in the zones.yaml file, but it seems that this was also stored to the DB file. is anything that I create added to the DB file?
Explained in there
Zones you create in the UI are stored alongside (almost) everything else you do in the UI - which isn't the database but some hidden files
but why is it stored in some hidden files if I configured zone ui: !include zones.yaml
Because ui means nothing
Literally nothing
The zone editor in the UI doesn't touch YAML
yes I know, I've stopped that container and I was able to update HA, but my question is why HA can't (won't ?) operate with another containers.
@humble mirage so the example on the bottom of https://www.home-assistant.io/docs/configuration/splitting_configuration/ is only for the automation part? for nothing else?
No, it's possible for all YAML
However, when you write zone ui: the ui part is just a label, it means nothing
You can write zone banana: and it has the same meaning
Because you’re running supervised. It’s a condition of running supervised. If you want to run other containers run HA container, then install all your add ons as docker containers instead
@humble mirage Ah ok, so it is just a lable to add different yaml files for some "area" of the main configuration, but it does not force the UI to store the configuration in that files?
@lone flint Discord isn't like IRC, you don't have to tag people on every response. Keep in mind that every time you tag somebody, they get a notification ping. That can very quickly become annoying and people may block you.
When using Discord's new Reply feature it defaults to pinging the person you reply to, click @ ON to @ OFF to stop this - on the right side of the compose bar.
You have to change this every time (thank the Discord devs for that).
Yes, it's just a tag, nothing fancy, no meaning at all - it's discarded. You can't force the UI to write YAML
It's a feature of YAML - explained in the YAML links I shared before
found an error when system restarted ```
Logger: homeassistant.setup
Source: /usr/local/lib/python3.9/asyncio/events.py:80
First occurred: 3:37:21 PM (1 occurrences)
Last logged: 3:37:21 PM
Setup of input_datetime is taking over 10 seconds.
thanks, now I understand
another one
Logger: homeassistant.components.recorder.util
Source: components/recorder/util.py:262
Integration: Recorder (documentation, issues)
First occurred: 3:36:58 PM (1 occurrences)
Last logged: 3:36:58 PM
The system could not validate that the sqlite3 database at //config/home-assistant_v2.db was shutdown cleanly
I fucked up - changed Host IP and can not access my homeassistant - Its installed in a VM Virtualbox - Is there anyway to access the backups to transfer to another instance?
@grand sun Not an expert, but might be worth mentioning which installation method did you use?
Folks, what is the consensus on when it's a good idea to update HA? Should I update to 2022.5 straight away, or wait a little while for bugs to be found and fixed?
This will be my first update since I set up HA 3/4 weeks ago.
I'm using the Container installation on a Pi 4
I installed with the .vdi file as Home Assistant Operating system .
I can only enter the HA CLI and not get access other ways atm.
Is there a way to change Core IP with HA CLI or another way of obtaining a backup inside the VM Virtualbox?
Looks like someone asked your question (if I've understood correctly) here:
https://community.home-assistant.io/t/how-to-change-ip-adresse-in-cli/332205/9
Maybe the answer in in there.
Sounds like login while in HA CLI will bring you to the OS context so that you can run other commands etc.
Some go instant, some wait for .3. It is up to you. You can roll back with container easily
Just make backups
Yeah I've started making backups now and rsync'ing them to my PC
I need to research how restoring a backup is done before I think, so that I'll know how to revert back quickly
With container you just copy the config folder. Restoring is as simple as copying it back
👍 I was just opening up a backup file to see what's in it. A tar within a zip within a tar 😆
Hi fellas. Can anyone guide me finalizing my first HA setup? It says my system is unhealthy due to a missing update on supervisor. It just won't let me upgrade it to the newest version.. is it a common problem and something I can fix myself?
Thanks in advance
Have you installed HAOS or HA Supervised?
But if you just started it should have been the latest version it installed
Exactly that. But it tries to update from 2022.04.0 to 2022.05.0
Ah right it was only just released yesterday so till before yesterday latest was 4.0
It should be telling you why it's unhealthy
Under system health it just states that Home assistant supervisor supervisor-2022.04.0 is unhealthy. When I try to update it says some json not found. But this is already far out of my comfort zone 😁
How did you install HA?
If you have a supervisor either you flashed an SD card or in a VM or you used a script
I flashed an SD card with a github file. I know nothing of scripts yet.
Ok then you have HAOS
If you go to Config --> Addons / Supervisor --> System it should be telling you there a reason for being unhealthy
in a red box (or at least have a more info button that tells you)
not finding it? 😅
Just lost all connection and I'm out of time, meh.
Bit Thank you for your time sir
Sound great. I will try tomorrow again
@lone flint posted a code wall, it is moved here --> https://hastebin.com/ubepesolot
@lone flint Rule #6: Please do not post codewalls (text longer than 15 lines) - use sites such as https://dpaste.org/ (pick YAML for the language), https://www.codepile.net/ (pick YAML for the language), or https://paste.debian.net/ (pick YAML for the language).
Please take the time now to review all of the rules and references in #rules.
I did that
and just got another zone called home, with correct settings but the "old" home (called Scruffy) is still there
or do I need to name the zone 'Home' instead of Home or home
Did you create the new zone in the UI?
I did create 2 additional zones in UI (nothing to do with the home zone/Scruffy zone)
and if I add zones via YAML file, they get added too, but the initial zone that was created immediately after the installation finished is still there
home has to be done in YAML - as mentioned in the fine manual
zone:
- name: Home
latitude: 0.0
longitude: 0.0
radius: 42
nope does not work either, it just creates a zone with the name Home
if I do it with Scruffy as name, same applies, just a second zone with the name Scruffy, but the initial Zone is still there
I also removed any other zone part of the configuration.yaml file and just added your part
I also restarted HAassOS after I added the part you pasted, same result, just a duplicate Zone
Interesting... looks like there's a bug been introduced recently
Is it just me, or has the map (in initial setup) where I could select my location, been removed?
I didn't see it when I set up a VM yesterday, it's possibly linked to the bug with the home zone
Want to see what issues are open (or were open and are now closed)?
- The core (backend) - for HA itself, and integrations
- The frontend (UI) - for cards and display issues
- The Supervisor
- HassOS
- The documentation
Don't forget to check for any alerts
If you think an upgrade broke something then:
- Downgrade to the previous version to check if it works correctly there
- If it does then change the
loggersettings for that integration todebug - Upgrade to the new version again
- Look in the logs for errors
- Open an issue, following the template provided and including the log information
I'd check to see if there's a known issue, and if not open one
thank you very much . my HassOS version (on a Pi3b+) is 2022.4.7 if that helps a bit
hi all, I need to completely reset my installation to onboarding state
I'm having a really hard time finding it. I restored a backup and now the system refuses connection to port 8123
Do you have SSH to look at the logs at least? Maye it hasn't finished restoring or something else is going on
I restored at onboarding and that's when the problems started. I think it was a version mismatch problem with the backup?
but I got no feedback from the UI and I don't know which log to examine
is port 4357 up?
also maybe you were restoring a backup where HA had SSL and you are still trying to access it over http:// ?
something like that, yeah. I'm thinking now reset to defaults and then manually add configs and add ons until I have my devices back
It's on a vhdx so that's not difficult
ah then you have direct access to the console of the VM
don't even need SSH
check first supervisor logs
yeah, the complete reset is working better. Lost a lot of customizations but I have the configs from the old tarball at least
you could also try uploading the backup from within Supervisor
I've often seen people having issues with uploading during onboarding
That sounds appealing
I hadn't updated my version of hassos for ages, the UI is a bit moved around
ooo that appears to have brought it all back
I'm not sure where I'm supposed to ask about update troubleshooting. I have OS installed and those channels are all archived. Is this the place?
Hi. I have just installed Home Assistant on a Raspberry Pi, added my Insteon integration (i'm using a 2245 hub) and it correctly located all of my insteon devices. However, whenever I try to control them the slider turns blue for about 10 seconds and they goes back to grey. None of my lights are coming on. If I manually turn a light on it does not show up as being on. Does anyone have any suggestions please?
Ok cool
I get an "unknown error, see supervisor" error when trying to update the OS. Any thoughts?
The logs don't mean anything to me, but I can post them.
@gaunt torrent there should be something in the supervisor logs about it
@little forum grey means unavailable so perhaps some networking issue. I'm not familiar at all with Insteon. Better to ask in #integrations-archived
This is what it says:
https://www.codepile.net/pile/KZGyn9A8
No idea what any of it means
Just moved into a new home, blank slate. Considering HA for my central automation platform, but hesitant about supporting my whole ecosystem from a Raspberry Pi due to SD card failures etc. I'm planning on wiring the home for ethernet and having a central run room with an equipment rack for patch panel, switch, modem, etc. Is my best option to slot a Linux server into the rack and run HA from that?
Or a vm if you have a server for that
Or see if you can find the upcoming Home Assistant Yellow
@high token just installed the 2022.5.1 and it all works, so yesterdays problem is solved. Thank you for your time
Does anyone know how to install HACS and or supervisor to HA core vers. on a synology nas ?
I checked advance use. but SSH dosent apper 😦
As long as you can SSH to the machine running the venv, just follow the docs
If you installed the venv using their software centre then you'll have to manually copy the files in
Thank you!
Is it possible on HAOS to dig further down on this 3743174 root 20 0 1681.6m 798.9m 73.7 5.0 145:41.41 S python3 -m homeassistant --config /config - the CPU use is the highest on the host now, but in the UI it say 0.5-1% load on the HA itself, and 25-35% on the core. Right now the mdns issue is stopped by removing free tagging on the HA port so the resolver doesnt hyper write to the log
that's the Core logs. You need the Supervisor logs
any sensitive information in that file or can I share it?
Should be fine, but you should also look at it
Can anyone take a look if anyone have time? https://drive.google.com/file/d/1Ie2dgpTZfbrz3n-lJYQ39seGxb3UoE6C/view?usp=sharing
When updating there are often warnings like "database is about to upgrade" and "adding index go database. this can take several minutes.. please be patient". What happens if you do a nother restart of the server right away while something like that is updating?
It shouldn't let you - though you can always force the issue and risk breaking things
Thanks
Supervisor didn't have anything relevant. I tried to install again this morning, went to supervisor logs and the last thing was from hours before
Ok thanks.
I'll have to wait until tonight to try that one.
you don't have access to the UI externally?
when an update starts the Supervisor logs should be saying that it's fetching the image (at the bottom)
UI yes. Not sure how to use terminal to update
you'd install an SSH addon
Ok. Haven't done that yet. I just use the UI extremely through the app. Save my troubleshooting for non work hours, haha
Hahaha
yeah it would be a pain to use the terminal from the app
you should definitely have that addon configured though and also so that you can connect over SSH from your computer
if HA goes down and the UI is not accessible, SSH is your way of checking what's wrong
restarting etc
I also noticed while poking around the updated ui that it says I have an unsupported install which is strange because I remember completely scrapping an old install a while back because it was unsupported and choosing a supported way.
it's HAOS isn't it?
flashed an image on an SD card or a VM?
in the Supervisor under System there should be a more info button to tell you the reason it marked your install unsupported
Yes, OS on VM.
I'll check
unless you updated to 2022.5 where the System stuff moved
but I assume the update error is general and not just to the OS
Oh. OS-Agent. I remember googling that before
System stuff moved?
Yea.. Again
System stuff is now under System 😄
Ah, didn’t really notice since I run HA on a k3s cluster
System, then System Health
Which runs fantastic by the way
What's a ova VM?
A vm format that can be imported into most vm software
I'm finding all sorts of new things in this layout
Stands for Open Virtualisation Appliance. As opposed to someone running HA in a Rpi 4 where it would say that instead
GM. trying to update HA OS and getting this error
22-05-06 10:45:06 ERROR (MainThread) [supervisor.os.manager] Home Assistant Operating System update failed with: Installation error: Failed marking slot kernel.0 non-bootable: Failed marking 'kernel.0' as bad: Failed to run barebox-state: Child process exited with code 1
trying to upgrade from 7.2 to 7.6. I also tried restoring HA back to last week and tried the upgrade again and getting the same error
I am running HA on proxmox if that helps
@slender sapphire Hmm not seen that before. I assume the same happens going from 7.2 to 7.3 ?
if you're in proxmox though might be quicker to spin up a new VM to restore that backup than trying to work out why
How can I go to 7.3. The install screen just gives you 7.2 to 7.6
7.3 of ... Home Assistant OS?
Yes
you could try from the terminal of one of the ssh addons
What's the correct command to set the ipv4 connection to static from the console?
net update enp2s0 --ipv4-address 192.168.1.250 --ipv4-method static returns an error: Can't update config on enp2e0: ipv4.addresses: this property cannot be empty for 'method=manual'
net info shows a v4 ip address, but it's wrong (192.168.1.240) and causing a clash on the router. I need to set it to static.
maybe you should make the Home Assistant with a static ip from your firewall
Static DHCP leases are good
I would love to, but the address it's clashing with is my server that's running unifi controller.
That sounds like a failing of your DHCP server, or you allocated the UniFi a static IP in the DHCP range
Yes absolutely. Static devices in the dhcp range is something that I need to fix. The problem was originally I left only 5 devices outside of dhcp and now they are all used up.
For now, being able to change the HA ip address from the console would fix my issue, but I'm getting an (erroneous?) error.
something like this
ha network update --ipv4-address 192.168.1.250/24 --ipv4-gateway 192.168.1.1 --ipv4-method static
I guessed your gateway though check the addresses
The gateway's correct. I'll try this in a sec. The misses has kicked me off the TV that I was using as a monitor. 😄
The command still fails with the same error message.
alternatively if you have a keyboard directly connected to it you can try with nmcli
I've rebooted the router to clear all the leases and i've assigned a static IP in the router.
I can ping homeassistant and 192.168.1.250 but get no webUI at 192.168.1.250:8123 or at homeassistant.local:8123
uugh, not a good day.
The ha console shows the correct ip: 192.168.1.250
core logs is showing an error in file awesomeversion.py. Probably the issue.
have you rebooted HA since setting a reserved IP and all ?
the awesomeversion package I think is associated with the .HA_VERSION file in /config
I prepared the SSD a few days ago, before 2022.5 was released. I wonder if that's the issue?
(this is running on a PC, not rpi)
in a VM?
No.
I thought you were running HAOS
I am, on a SSD in a micro-pc
No I decided not to do that at this point as i have no other use for this PC at the moment, maybe in the future.
cat /mnt/data/supervisor/homeassistant/.HA_VERSION
How do i exit ha> to a normal cli ?
login
do a rm of that file
rm /mnt/data/supervisor/homeassistant/.HA_VERSION
?
yeah
ok, done.
ha core restart
Unknown error, see supervisor
ha su logs
Home Assistant has Crashed!
There are a load of warnings too about failes to initialize by udev
Seems to be unused usb ports, maybe.
Looking through the log, it looks like HA crashed, did a rebuild, then crashed again, then crashed again.
hmm
try ha su repair first
assuming it finishes then try ha core rebuild
although it has tried rebuilding itself already
Its a fresh install. Is it worth messing about, I can just pull the drive and re-install it?
to be fair that could be faster at this point
rebuild failed.
try one last host reboot just in case haha
one other possibility perhaps if you're going to reflash
rm /mnt/data/supervisor/homeassistant/*
then ha host reboot
that should delete every single file in HA Core
and trigger a new onboarding
had to add -r to that command. It's rebooting now..
ah yeah there were folders in there right
json decoder.
and helpers.storage
ok, reflashed the SSD and i'm at the "preparing home assistant" screen 😄
way quicker 😄
and now has a reserved ip from the router right?
yes
perfect
My backup will set it to static also as i'm using adgaurd.
I'm moving from rpi to PC.
backups don't include networking setitngs though
ah ok. I'll go in and change that again later.
Gonna take a while to navigate this new gui !
😄 Not in add-ons any more then.
My backup was from 2022.4.7
Should I do a partial restore of only folders, or do a full restore then update again to 2022.5 ?
same thing at this point I think
Woo, i'm back up and running on a microPC. Amazing.
Hi, is there any chance to get official checksum (or GPG/PGP signs) for files downloaded from GitHub - Releases - Assets?
I just installed home assistant for the first time and trying to get my reverse proxy working correctly. could somebody help me with this?
You need to set the XFF and trusted proxies entries in http:
# Example configuration.yaml entry
http:
use_x_forwarded_for: true
trusted_proxies:
- 172.30.33.0/24
``` (for example)
http:
use_x_forwarded_for: true
trusted_proxies:
- 192.168.1.0/24
To format your text as code, enter three backticks on the first line, press Enter for a new line, paste your code, press Enter again for another new line, and lastly three more backticks. Here's an example
Don't forget you can edit your post rather than repeatedly posting the same thing.
For over 15 lines you must use a code share site such as https://dpaste.org/ (pick YAML for the language), https://www.codepile.net/ (pick YAML for the language), or https://paste.debian.net/ (pick YAML for the language).
you probably don't need to open your whole network
thats what I have so far. and I can connect to it, but when I try to log in, I get this error in the log: Login attempt or request with invalid authentication from pfSense.example.com (192.168.1.1). (Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/100.0.4896.127 Safari/537.36)
RobC yea I know, but I was troubleshooting and switched it to that for now.
Is the proxy configured to support websockets?
I don't get it - I did a third install on a third SD card on a raspberry pi 4. Pattern repeated itself. Everything works fine and then, this time after only 48 hours it just stops working. I'm using an official power supply and a scandisk micro SD card, that should be fine for the job. Apparently not. I really don't get it. Why would HA even suggest to put on a Raspberry Pi if all it does is fry the SD card?!
WebSockets Support is enabled.
It shouldn't fry SD cards quickly, unless they're already on their way out or low quality (minimal write life). However SD cards die eventually when used in a Pi, it's purely when - and Home Assistant itself is hard on them.
Of course, if you can re-flash the same SD card and keep going then it's not killing the card
I haven't tried to flash other things than HA on again - but that doesn't work 🤷
The Pi is, historically, cheap and easy to get a hold of, which is why it's been recommended. It is an educational toy though, and has some significant shortcomings.
Get yourself a microPC and a SSD. The difference between a pi and a pc is night and day!
and they cost the same. (ebay)
Yes, a optiplex 3040 is £95 with 8G ram and a 120G ssd from ebay. The new Pi4 costs the same.
A 3040 is over-powered for just HA. A cheaper 3020 would do.
I did just that thanks to a recommendation from Tinkerer a few days ago and a video on YT by digiblurDIY compared a Pi and a PC.
Well just frying SD cards is pretty much spoiling the fun I was hoping to have automating stuff. So I'm very much open to suggestions 😄
I found a used optiplex 3040 with 8gb ram and 240 gb ssd for around 65 EUR
Is there any specific reason youre recommending the optiplex?
No, just what I bought.
As long as it has at least 1 normal (not USB3) port if you use zigbee stick, it'll work.
I bought a micro pc version.
6100T processor.
Greetings all - I'm trying to sort out the best way for me to install Home Assistant. (Silly question) I would like to run my Ubiquity cloud key from the same Raspberry Pi 4 that is hosting my HA installation. Can I install other containers if I chose HAOS/Supervised OS install? Or would it make more sense for me to manually install Docker and install both the Core and the Supervisor. I guess the third option would be to manually maintain any addons and skip the supervisor? Sometimes having choices is a bad thing 🙂 Thank you for any suggestions in advance.
Can I install other containers if I chose HAOS/Supervised OS install?
No
you should use a container installation for HA if want to run other things from your Pi
Thanks @grand pivot I'm still rather a Noob with HA. If I install just the container, can I still integrate add-ons? I'm thinking of things like Node-Red.
Okay, so I have to manage the integrations by hand then. That makes sense.
Okay, thank you!
addons are just other software, and you would need to install them yourself and do whatever configuration is required
That makes complete sense.
Off-topic but, Is cloud key the same as unifi controller?
@lime cobalt the controller is the correct term. Cloud Key is the hardware version, but I couldn't recall the name of the software.
There's a "unifi network application" add-on in the add-on store. not sure if that's the same.
I use this and it works great: https://hub.docker.com/r/jacobalberty/unifi
I have been using the version from Jacobalberty as well, How do you host the service? I had it working for a while with macvlan, but when I attempted to upgrade, it bombed.
docker-compose on my Linux box, the same one that hosts my HA instance in a KVM
nods did you do anything special for the networking on the UI container or just vanilla?
I used the example docker-compose.yaml pretty much as-is: https://github.com/jacobalberty/unifi-docker/blob/master/docker-compose.yml
gotcha. Thanks!
@empty river I've just installed the HA Unifi add-on and that also works. So you have both options. Docker or HAOS.
@lime cobalt Thanks for that info as well!
Hi all, I have the following error whilst tryinng to update through SSH
local record: tls: bad record MAC
Hi all, new-HA-starter here. My new installation of home assistant to my Raspberry Pi 4 Model B appears to end at the 'Welcome to the Home Assistant command line." It says "Waiting for Supervisor to startup...
System information
Info listed:
IPs addresses
OS Version
Home assistant Core
URLs
Am I doing something wrong?
You should focus on the web page at http://<youripaddress>:8123 rather than the command line
Thanks
how can i fix error 400? i tryed everything but its not letting me connect to the outside of my network
Hi all, I'm running homeassistant os 7.6 in a VM with HA 2022.5.2 but am struggling to set up a static IP. I've tried through the web UI and the CLI but get the same message: Error: Can't update config on eth0: ipv4.addresses: this property cannot be empty for 'method=manual' does anyone know what I'm doing wrong?
(I used ha network update eth0 --ipv4-address 192.168.1.93 --ipv4-method static --ipv4-gateway 192.168.1.1) on the CLI
nevermind, I managed to get it to work via a root terminal
Just saw the message, also have the same issue!
anyone know a really easy way to update container installations using docker compose? or does it always have to be remove stack -> get new image -> readd stack
just slightly annoying to do the entire process
docker-compose pull
docker-compose up -d
ah
I could use some help with setting up Home Assistant with Hyper-V. It was working, but suddenly stopped getting an IP address.
thanks
Home Assistant makes my ESXi Crash
I can't figure why Home Assistant makes my ESXi crash.
When Home Assistant makes an update, it stops working and I need to restart my server.
All other VMs come back online, but not the Home Assistant VM, the vmdk file is corrupted.
If I reinstall Home Assistant with my last full backup it works until the next update.
Does anyone have some tips on how to resolve this issue?
Thank you in advance.
Don't know if my issue can be posted here.
Has anyone had experience using HA on a synology NAS?
Depends on the NAS, ideally you want one that can run VMs
Is this the best channel for support related to upgrading core? Issue possibly related to Supervisor....trying to pinpoint, really.
Specifically it seems to hang during update process. Logs are unavailable (HA hangs / times out waiting for logs)
Even trying to view logs through SSH does the same thing:
Usually updates take 5 minutes or so; but this just hangs for 20 minutes and I have to power cycle my Pi - Observer shows no issues; however
I am not really sure where to start with diagnosing this
i have some screenshots; but looks like I cannot attach them here
Please use imgur or other image sharing web sites, and share the link here.
Image posting is blocked in most channels to discourage people from sharing text as images. Sharing text as images assumes that everybody sees the world as you do, which isn't the case. Some people are colour blind, or have visual impairment that means they can't make sense of an image of text.
someone care to share their docker-compose stack? need ideas on what to and how to include
@fresh marten include what?
Here is my kubernetes stack if it helps a little https://github.com/miwagner1/k8s-at-home-with-k3s/tree/main/stacks/homeassistant
Probably mqtt and an external database for most, plus a reverse proxy
😄 I should start with kubernetes 😄
I rather mean HA mosqitto, grafana, infux
And WireGuard if you don’t want to expose it
oh yea, I should get vscode again
I need to learn how to use variables with kustomize so I can just have my git mirror my current setup
I’m trying to make the kubernetes entry level lower
shouldn't you rather use secrets?
But fill them out with an env file that is ignored from being uploaded to git
I have a lot more running… but I’m too embarrassed to upload them yet ;p
😄 i know this
But yah, technically Home Assistant has 2 ips it can be accessed from, the node it’s currently running on since it’s in host mode to allow mdns and everything to work including HomeKit, and my ingress ip from my reverse proxy which means it does not matter what node it runs on
I hope at some point I can get good enough to make a HELM chart to be added officially and manage it
I have home assistant container, is there such a thing as just adding a supervisor container next to it so that all the addons and store or the fancier features do work?
no
is it possible to add supervisor in some way to that setup? I don't want an image OS and I like docker.
you just asked that. no
if you're accustomed to using Docker containers, then you can install whatever you want that way
it wasn't the same question, if you read closely you'd see there was a difference.
the answer is the same
there's no nuance there
"add supervisor" and "add something like supervisor" are the same thing when the goal is to use addons
You could switch to home assistant supervised but there is no way just to add the supervisor to an existing container install
thanks. I couldn't find a description of the structure of that option. the instructions list requirements and then a package to install, but no description of what a system with that ends up looking like. and of course it says, a requirement is "OS is dedicated to running HAS" .. which is of course not what I'm planning to do, so ..
Well if you run it alongside anything else it is concerned unsupported and your mileage may vary as to if it will work
I was concerned about this quote: "In case any abnormality is detected that prevents Home Assistant from functioning, the Home Assistant Supervisor will report this to the user and block updates to prevent installations from breaking." which sort of sounds like if additional software is present it's going to freak out? seeing as how the req are listed as "conditions". But I guess probably not.
Hi folks I am new here and I am trying to move my install from a raspberry to a VM all went well and worked until I needed to do a firmware update then I received this message NFO Reading configuration /config/esphome/water-tank-sensor.yaml...
INFO Generating C++ source...
INFO Compiling app...
Processing waterlevelsensor (board: nodemcuv2; framework: arduino; platform: platformio/espressif8266 @ 3.2.0)
HARDWARE: ESP8266 80MHz, 80KB RAM, 4MB Flash
LDF: Library Dependency Finder -> https://bit.ly/configure-pio-ldf
Dependency Graph
|-- <ESP8266WiFi> 1.0
|-- <ESP8266mDNS> 1.2
Compiling /data/waterlevelsensor/.pioenvs/waterlevelsensor/src/main.cpp.o
Linking /data/waterlevelsensor/.pioenvs/waterlevelsensor/firmware.elf
*** [checkprogsize] xtensa-lx106-elf-size: Permission denied
========================= [FAILED] Took 10.33 seconds =========================
Yeah
I cannot figure out how to manage directories
Honestly your best bet is a VM
I'm running on a pi4 which I put a significant amount of effort into consolidating various functions into 🙂
currently works great, but maybe I need to separate HA off onto a pi3 or something
Got it. In that case just stick to container till you can dedicate a device to HAOS
Access denied in VM
I did misinterpet that quote I think? HA supervised won't shut down or stop updates just because it detects an unsupported condition, it's just saying that if it starts crashing or thinks it is broken, it will stop?
From what I know it just blocks updates. But I haven’t tested it
I guess I'll have to forego the convenience of supervised/addons/etc for now
I found this on the forums. I don't think I need supervised in my life with this level of invasiveness. (Oh, no inline pics it seems. https://community.home-assistant.io/t/homeassistant-hassio-supervisor-in-docker/393415/3 )
Fair enough, it all depends on what you plan on running
thanks for the info
If you really want HA os, I would keep to the official addons and perhaps one or 2 others. Some addons for whatever reason cause the entire os to eat dirt
that really shouldn't ever happen, what addon were you using that caused the OS "eat dirt"
How can I install a supervised docker image on my existing linux installation?
Please don't. There are installation instructions with the rest on the HA site if you must
I saw instructions for installing with docker but it doesn't install supervisor? (it's rpi)
Correct
so what's "the rest"?
If you want the supervisor, Home Assistant Operating System image for pi.
was it removed? I already had it before I got rid of the install I had
Was what removed?
docker with supervisor on rpi
It's all on the installation page
But do yourself a favor and pick a different installation method
A Supervised Install is a terrible way to install HA. As Rob said, please don't do it. If you want the supervisor use a HAOS image or HAOS running in a VM. I have no idea if the pi supports VMs though.
Cool, guess I'll just figure it out myself
Anyone know why Google account linking has stopped my local access to HA? I can connect when I’m not on the local WiFi network, but my laptop and phone can’t connect after attempting to integrate Google. Google Home also can’t connect.
does anyone know what these errors are being caused by its a virtual machine https://imgur.com/a/vY46UJP
I have a 2X ESXi Servers with VSAN/VCENTER running downstairs anyways. As soon as someone suggested to use the HA OS image to build a VM, I got everything up and running in less than 20 minutes.
Does anyone know if HA has released a disk image that could be used to build a VM in Unraid or other similar "free" hypervisors?
@deep edge there are several VM images built. I believe the KVM image works with Unraid
the kvm one runs fine with proxmox
Hi all, struggling with something that I can’t work out… I have ha running on a windows pc under VMware. It’s an old version and can’t update due to loads of configuration errors, so want to start afresh with a new install, have downloaded new version .vdi file, and followed the setup instructions but when I start the vm it never gets an ip address, I have everything setup the same as the 1st instance, any ideas …. Thanks in advance
You've got it using bridged networking with the same adapter?
Yes bridged and same adapter… I thought that by shutting down 1st version the 2nd would be ok, is this not the case?
Shouldn't matter, both should get an IP from your DHCP server
That’s what I thought… lcant see why it’s not getting one
Can always delete the /config and work through the errors to update to current version 😅
Was planning on,leaving that one running whilst I got a clean(er) version running without all my mistakes, copied configs etc… hence wanting to have a second vm running. Have check with Fing network tool and it is seeing homeassistant but it’s not got an IP address, think I will check in on my unifi setup next…definitely something weird going on
could also try a setting a static IP on the console
Is that via the con edit - I vaguely remember doing on my 1st version
nmcli yeah
ignore the bit about wifi though. it has also instructions on static IP (and you need login to escape the ha cli)
Thanks will try this
Could have been frigate, all I know is that the pi4 would perhaps last a week before going derp
Same pi has been up now for a couple weeks on my k3s cluster
Can’t buy any more pi4s for further testing since they are all out of stock and the ones I have are in my Kubernetes cluster
I really want at least 1 more :p
I have been trying to install home assistant on my raspberry pi 3 b for 3 days but it always gives me the same error I thought it was due to lack of space (it is what I have found on the internet related to the error) I put a 32g card but I bought one 64g class 2 as recommended on the web but nothing I continue with the same error, I have tried installing both the 32-bit image and the 64-bit image I have tried with balena echer and with rapberry imager but nothing same result I do not understand what it is what happen. If anything comes to mind it would help me a lot. thank you https://prnt.sc/AZMac-Q4F43g
I've no experience with what you're attempting to do, but here's a few ideas:
1.It is complaining about not being to able fetch some files, does it have unhindered access to the internet?
2.There's a bunch of 404(not found) errors, is the software you're using recent? If you're following some documentation, direct links to specific versions might not be relevant anymore.
1- the raspi is directly connected to the cable router.
2- I am following the steps of the official website https://www.home-assistant.io/installation/raspberrypi
Ahh, sounds very reasonable, sorry, I have no raspberry PI to test it from my network. If you can ssh in to your raspberry and do some successful network tests, I do not think the problem is on your end.
Also, OCR-ed and googled your issue, unless there is a great mishap in coding, a "404 docker image not found" error would not be in any way related to local storage, sounds very misleading.
How can removing everything from /config not trigger a new onboarding? But instead result with same issues on .storage files being borked and all dependencies erroring out?
I've completely run out of ideas
@weary patio posted a code wall, it is moved here --> https://hastebin.com/ifilegojak
too much info in one post
this isn't the forum 😅
if you have a keyboard and monitor connected you'd want to look at the logs (ha core logs and ha su logs)
interestingly have just been dealing with another user on the forum with an Odroid N2 they just bought and having issues
(unless you are that user haha)
Hello all, HA noob (Insteon exile) here for the first time, hoping someone can help me connect to my Home Assistant Bundle - Standard ODROID-N2+ 2GB/16GB eMMC. Server wasn’t found when Scanning for Servers in the iOS HA app so I connected a keyboard and monitor and obtained IP address. When I type http://192.168.1.107:8123/ (or https)
In a browser I get the message 192.168.1.107 refused to connect ERR_CONNECTION_REFUSED. Following another suggestion in this thread I tried http://192.168.1.107: 4357/ and I get this message:
Home Assistant observer
Supervisor: Connected
Supported: Supported
Healthy: Healthy
Tips/pointers greatly appreciated!
how do you use hassIO if you don't have raspberry? because right now I'm thinking of changing to something else XD
Nope, different noob but we seem to be having same issue...
There are several ways to run Home Assistant OS without using a RPi
Well, the original guy might have more insight, but I would just always prefer a linux box/container if you are looking for options 😄
@weary patio are you familiar with nmap? IMHO, if your router isn't interfering, sounds like whatever pre-built solution you are are using is broken.
Yah, I'm wondering same thing. I'll ping the supplier.
with virtualmachine? https://www.home-assistant.io/installation/linux
Yeah, if you would like to tinker, (no idea how yours is set up). I'd look at some logs, for mine it would be something like journalctl -u home-assistant@homeassistant.service -f to watch the log in realtime. Clearly the HA installation doesn't seem healthy.
connect a monitor and keyboard and run ha su logs first then if nothing looks wrong there ha core logs
Sounds like an ok option, but keep in mind that your hypervisor(presumably your desktop PC) should be on 24/7, which might be less than ideal for most.
Got it. The log doesn't seem too informative; this part seems most relevant and I'm not sure it's telling me anything I don't already know... unless I'm missing something? "LocalDataTask <7B621AC1-A610-4D9F-B185-0CEEC8F5E15A>.<1>"
), NSLocalizedDescription=Could not connect to the server., NSErrorFailingURLStringKey=http://192.168.1.107:8123/auth/authorize?response_type=code&client_id=https://home-assistant.io/iOS&redirect_uri=homeassistant://auth-callback, NSErrorFailingURLKey=http://192.168.1.107:8123/auth/authorize?response_type=code&client_id=https://home-assistant.io/iOS&redirect_uri=homeassistant://auth-callback, _kCFStreamErrorDomainKey=1}))
@ripe fossil Just reiterating, but some kind of small, power efficient computer running the whole thing seems to be a popular choice
@weary patio no idea b/co the setup, but you might try restarting it and finding some more relevant stuff in the logs, why it does not want to listen on port 8123
I'll give that a shot; strangely it does appear you can connect to port 4357:
Home Assistant observer
Supervisor: Connected
Supported: Supported
Healthy: Healthy
Doesn't k8s/k3s require a lot of ekstra resources? And also what in your mind is the main reason you chose to deploy HA with k8s?
About 550 MB on the master nodes if they are also workers (default)
50 MB on worker only nodes
ram that is?
Yes
Because I am using k8s for everything? A node can fail and HA will migrate over if it was on the one that failed/updated/need to be rebooted
@weary patio I came here today having the same issue. I can reach the observer on port 4357, but I cannot reach 8123. This was all working last night when I was setting up some automations. I have even isolated the pi off my network using a switch to connect to my laptop.... same results.
hollup - are you running it on a pi-cluster?
how many pies?
yeah, I guess that just requires an ingress yeah?
What are you using for storage?
Nope, I could give HA itself it’s own IP like I did with wireguard
Longhorn
I'm going to toss the Odroid in the trash and try using a different platform. Was hoping the unit from AmeriDroid would be plug-and-play but this is torture. If anyone has a suggestion for a platform for us non-technical Insteon noobs I'd love to hear about it - TIA!
And what hardware?
It seems that a lot of ppl are experiencing issues with HA just imploding on itself on single board computers
(myself included 😅)
ooohhhhh clever
Did you use Jeff Gerlings design?
And do you run things on the cluster that aren't related to HA?
My own design, I’m learning as I go and yes
See my repro
I’m adding a lot more
if you could share pictures of both logs we might see something relevant. This specific thing not sure what that's saying
Hah, cool
Tanoshi... everything really do exist on the internet 😅
I have a lot running at home, just haven’t published them yet as they are not up to my new standards yet
scrypted looks nice - whats your experience with using it?
Im a mod on their discord, it’s super nice
Authelia needs a redis database for logon persistence
Lots to do
Http3 support with traefik is done, just need to test
:p
Camera
@high token - I am also getting the connection refused in port 8123. Here's both logs.
https://snipboard.io/MhwCfO.jpg
https://snipboard.io/kEamzA.jpg
it's an Odroid so likely eMMC
the core logs aren't really saying much apart from the exit code that I don't know what it means
if you don't care about the warranty I'd try reflashing the eMMC
[info] Start Home Assistant Supervised
[info] Installing the 'ha' cli
[info] Within a few minutes you will be able to reach Home Assistant at:
[info] http://homeassistant.local:8123 or using the IP address of your
[info] machine: http://192.168.178.90:8123
How long will this take
I'm on a pi 3b with install on SD card
A few minutes 😊
Start Home Assistant Supervised
Well my experience - and i might have bought a bad batch of sd cards is that home assistant will simply destroy them in no time
what is xd?
Supervised... is a trap
Most people who run it ignore the requirements, assuming they read them, and then bitch about a broken system
It gives you nothing you don't get with Home Assistant OS, other than having to manage the OS yourself
i want to make more with my pi so i cant install this os ;D
Then you're going to ignore the requirements... called it
https://github.com/home-assistant/architecture/blob/master/adr/0014-home-assistant-supervised.md
No additional software, outside of the Home Assistant ecosystem, is installed
I'm still able to see that HA is up and running to some extent when I connect monitor and keyboard. With a damaged card, wouldn't it be dead?
It probably will be in some time
You'll be back here complaining that the Supervisor is reporting that the system is unhealthy and unsupported, and that you can't upgrade...

If you want to "do more" with the Pi then install Home Assistant Container - you don't get add-ons, but you do get freedom to do what you want
Another poor soul; #installation-archived message
Another person with insufficient pain and frustration in their lives 

Here's a question. If you delete .storage entirely (and of course having defaults on the yaml files) the files in .storage should be recreated shouldn't they? And presumably trigger an onboarding or?
If you delete that and not configuration.yaml I'm not sure what happens
I'd hope it'd trigger onboarding, but I wonder if that's an edge case
configuration.yaml really just having default_config only
I've been trying to help this guy on the forum. Freshly bought ondroid n2 with what seems to be broken .storage
first had him delete everything in /config
but didn't triiger onboarding
instead empty files in .storage, which then causes a bunch of dependency errors and crashes core
I've run completely out of ideas, apart from reflashing
404 is an HTTP status code error
okey here is my problem:
22-05-08 21:54:07 WARNING (MainThread) [supervisor.jobs] 'Updater.fetch_data' blocked from execution, no supervisor internet connection
22-05-08 21:54:07 WARNING (MainThread) [supervisor.plugins.cli] Error on install cli plugin. Retry in 30sec
you have any idea why there is no "internet connection" i checked all requirements and all looking fine ;D
@humble mirage I think I'm gonna go with installing on a used laptop as you recommended. How much storage would you recommend getting?
More than 64 GB
How come it needs that much?
was thinking about a thin client - but i guess thats ruled out 😅
32 GB is the basic minimum recommendation in the docs, but some people still run out of space when they load a bunch of add-ons or want to keep a year of history
Damnnn I just succeeded in getting HA up and running on my synology nas through docker only to find it has no store :/
Add-ons are just software running in Docker anyway
As long as ... what?
the docker containers
In the same network or otherwise accessible
Basic networking... routing of packets and all that jazz
Is it possible to get an overview of the add-on docker images somewhere?
What do you mean... overview?
well i don't have the store anymore
i'd still like to know the docker registry name for that duckdns add on
There may not be a 1:1 mapping
Some add-ons may be custom built, not based on an existing image
https://docs.linuxserver.io/images/docker-duckdns and https://docs.linuxserver.io/general/swag will give you all you need instead
So basically HA in docker is HA in hard mode 😄
Nah, hard mode is Core. Pain mode is Supervised
If you want add-ons you want to run HAOS in a VM
Just trying to move from OS to container, can it be that the backup integration does not have a restore feature or am I just to blind? (That's a real possibility)
Although I cannot verify the latter, the integration can backup a container/core installation to restore in a HAOS/Supervised installation
you need the Supervisor in order to restart a backup
you would just unzip the backup and move /config over to get everything back (minus the addon stuff obviously)
eeh, bummer. lets see if i can just untar it and then place the files in the config dir.
just stay in HAOS 😛
the install needs a refresh xD
Somehow HAOS just got stuck on multiple occasions and wasn't able to reboot by ACPI. wife approval -25 points
I'm reliant on these points
with container I can at least dig out what broke (at least in theory )
also, someone has a clue why the container needs the privileged flag?
I don't think it's necessary if you know what you're doing
I wanted to see if there are some other ideas on what the problem is or at least the least painful way of getting back operational.
I am running HAOS on rpi 3B for about a month (Insteon refugee).
As of today only I cannot access HA via port 8123 (ERR_CONNECTION_REFUSED). I can see the device on my router and can ping successfully. I can also reach port 4357.
I can connect monitor and keyboard and get to HA CLI. Logs are posted above.
I was able to put only the pi and my laptop on a wired network using a switch. Same results as above.
Last night I added a mobile device to track, add some automations, and reserved the IPv4 address in my router. I was able to access fine last night before bed. I did not have external access or any port forwarding rules turned on.
I have a few recent partial backups from the May releases and were not downloaded, my only full backup was just after initial config and was downloaded.
Several of my sensors seem to have gone from F to C, but I'm still set to imperial on the settings. Any ideas on how to fix this?
use metric lol
i'm trying to migrate a container installation, but I get an onboarding screen instead of a login, as though my existing accounts are not recognized. i went so far as pullin my full folder as root with rsync, but still getting the 'create user' screen. any ideas as to get my to show up?
Did you check that all of .storage/ was copied?
The only other options are either that you didn't point the new container at your config folder, or that you're not running as root for some reason and it's permissions
folder is right, at the very least, as it's /docker/homeassistant 😅
i've got the same files in there
follow up on failure/success. did not copy all the files the first go, then mixed up command order and replaced my original (prod) files with default files
backups 🙏
That doesn't answer my question. Automations that used those numbers are now all broken, as well as dashboard settings for gauges.
weird migration question on docker. got HA and nodered up and running, but my history/graphs on my phone match the old one, even though it isn't running, and i have the new ip set up on the app. more of a curiosity, but i'm baffled as to why the graphs (24h temperature, humidity, etc) match the old iteration.
Well, if you migrated HA I assume you migrated the database too?
right, but my phone and web browser don't match, pointing to the same instance
i can tell because there is a gap where it wasn't running for an hour, or whatever
where the old one didn't have that gap
Hello friends! I have a HA networking-related question to ask.
TL;DR is that the installation of HA on a Raspberry Pi4 is successful for me. Whenever I don't use the internet provider's router (which I must use for a 2Gig connection). I can connect to the Home Assistant web front. However, whenever I introduce the router, I am no longer able to connect. I understand thT the router is the ulprit but I don't have the necessary networking expertise to realze what needs to be done to fix the situation. Any help and advice here are greatly appreciated!
Are you trying to access the instance via a dns name like "http://homeassistant:8123" or http://homeassistant.local:8123 ?
If so one thing worth trying is to reboot any device trying to connect to home assistant, as it may have cached the dns result from the other router, and the two different router probably assign different IP addresses. If that does not help it is possible the ISP provided router is not providing dns resolving in the way you need, and/or is blocking mdns discovery. You may need to try out IP address based access, at least to confirm this. Then it is probably worth looking through the router settings to see if something is configured screwy.
If however, you are trying to connect via an IP address, and that is not working with your ISP provided router, then you probably need to find and turn off some form of "client isolation mode". once Be aware that your ISP router is likely assigning a different IP address than your other router.
hi all, I have a Pi4 running home assistant OS. does the OS take care of upgrading the Raspberry firmware or should I do it manually?
Hi all, is there a way to run home assistant using Nabu casa with no local instance
No. Nabu casa just interfaces with your local installation
Possible to run the local instance also on a cloud server instead?
I noticed that tuya devices anyways sync with the tuya cloud
What’s the point. The point for HA is local control. Moving it out means any devices that are local (eg if you got zigbee stuff & a network zigbee stick) you would have to tunnel to the remote instance. It’s not worth it
If you already use Tuya cloud stuff then having HA in a cloud instance will just make it more complex and prone to failure
It's actually a very interesting project. I want tuya button click to call a custom API. But i don't want to host a device which may have internet issues. I'd rather have a cloud server since it guarantees internet
Dont tell this AWS or the place I work at 😅
Yes, you can see it in the release changelog that it's included
thank you very much 🙂
Did you actually read that before you posted it???
Sorry I'll try to explain better. As you would know tuya doesn't allow me to call APIs of other platforms. I am trying to use home assistant as an intermediary to detect the click of a tuya button and be able to call a custom API. I want to avoid using a local instance for this as have a cloud instance since the api i am trying to call is on a cloud.
Did i explain better? I hope ...
I got it the first time, but if there's an internet issue then the device isn't going to reach it's own cloud, much less your cloud server.
Hi! I'm hoping someone can provide me with an idea of how to resolve this. I've been through the forum and through the FAQ, can't seem to find the fix. I just flashed a fresh install using BalenaEtcher. Using a Raspberry pi 3 B+. Also tried switching out the Pi, and switching out the network cable, and the port on the router. For whatever reason I'm not seeing the device on my network. Using a wired connection on a brand new fresh install.... haven't even made it to the setup part.... What am i doing wrong?
What image did you use on balena etcher? The pi3 one?
yes
I've tried the 32bit, and the 64bit.... by all "lights" there's traffic, but my router doesn't display it as one of the connected devices. I'm at a complete loss what's going on and what to do.
Have you tried just doing the home assistant.Local address?
you mean homeassistant.local:8123?
Yes
does not resolve
SOLUTION: ameriDroid ODROID-N2+ refuses to connect to your network: Reflash the eMMC. Warning: I’m NOT a tech geek so this was painful for me but I am persistent and so I typed up how to do this hoping it will help other non-techie Insteon exiles out there and posted it to the forum: https://community.home-assistant.io/t/new-odroid-n2-with-preinstalled-os-wont-load-ip-8123/416034/2. HTH!
ameriDroid ODROID-N2+
hi, im stuck at preparing home assistant for more than 1 hour, the error say that Not Found ("no such image: ghcr.io/home-assistant/raspberrypi4-64-homeassistant:2022.5.3: No such image: ghcr.io/home-assistant/raspberrypi4-64-homeassistant:2022.5.3"
can someone help me
Undoubtedly some network or DNS issue
tq sir
guys can someone help me quickly with my hass ? looks like it's not starting again and I can't login with my user on the webgui
Do you have any logs or anything? Or just it’s black
forget it, I was trying to fix it from the CLI I think I did core rebuild and looks like this has reset my home assistant from scratch
I have now the prompt to create the user as the initial setup 😦
don't know what happened to be honest - it's my first setup with home assistant and I feel like using the OS makes me blind as I can't access to the system
Then installing Debian/ whatever Linux server OS and running container is probably your best bet
I think it's what I'm going to do
I don't know usually what's people do ? are you using the os version ? or self-installed ?
core version or Supervised ?
I run container. Its the docker install (so using docker compose). Not supervised or core
any limitation in term of add-ons ?
With container you run whatever other containers you want. All add ons are essentially containers so you can run them yourself. Just there isn’t a managed way to run it
@primal nimbus did you delete anything? core rebuild would not reset your config
And there's nothing for you to do at the OS level, anyway. Everything is in containers and the data in volumes
how could I delete something ? from what I saw I don't have access to anything on the OS.
So I did try some stuff like
- supervisor restart / repair / logs
- core restart / rebuild /logs
I saw some logs about something wrong in the config of hass
I rebooted another time the VM and the service came back reseted
can't tell more tbh - it's not a big deal I'm at the beginning with hass 🙂 so nothing was done deeply in it yet
Side note, this is why we have backups and why they’re kept off of HA
yep I will 🙂 for sure
So did it let you create a new user and you could log in with it?
yes like a fresh new install
well I probably done something wrong somewhere 🙂 as a big hass noob
ah nice thanks I knew the studio code but not the Samba addon
Hi is there someone that could help me 1 sec?
I just tried instaling hass on a rpi
if you explain what you need, perhaps
File "/srv/homeassistant/bin/hass", line 10, in <module>
sys.exit(main())
File "/srv/homeassistant/lib/python3.7/site-packages/homeassistant/__main__.py", line 278, in main
args = get_arguments()
File "/srv/homeassistant/lib/python3.7/site-packages/homeassistant/__main__.py", line 60, in get_arguments
import homeassistant.config as config_util
File "/srv/homeassistant/lib/python3.7/site-packages/homeassistant/config.py", line 52, in <module>
import homeassistant.helpers.config_validation as cv
File "/srv/homeassistant/lib/python3.7/site-packages/homeassistant/helpers/config_validation.py", line 85, in <module>
from homeassistant.helpers import (
File "/srv/homeassistant/lib/python3.7/site-packages/homeassistant/helpers/script_variables.py", line 6, in <module>
from . import template
File "/srv/homeassistant/lib/python3.7/site-packages/homeassistant/helpers/template.py", line 19, in <module>
from jinja2 import contextfilter, contextfunction
ImportError: cannot import name 'contextfilter' from 'jinja2' (/srv/homeassistant/lib/python3.7/site-packages/jinja2/__init__.py)```
it crashes with this on startup
did everything command by command from the website
I tried to install home assistant OS on an old laptop, i used an ubuntu 20 to flash the os with balena and when everything finish i restart the laptop and says "No bootable device -- insert boot disk and press any key"
sounds like your old laptop may not support EFI boot, or it's not enabled
I entered again the live ubuntu and use cfdisk on the disk and all the partitions were created so i dont know
Could be its old asf its a hp mini 110
I search the efi boot in the BIOS but i didnt found
When did HP Mini 110 come out?
September 24, 2009
It went on sale on HP's online store on September 24, 2009 for $399.99. The laptop can be customized with either the Intel Atom N270 or N280 and uses the nVidia GeForce 9400M G graphics used in the ION platform.
So you think its possible?
I think it's probably not possible
Those kind of things can be hass trying to download a python package but not getting it in time before it's needed. Give it a few minutes to complete that and then just restart.
Also, for installing on an rpi I highly recommend going the HAOS route.
Ok thanks
I had to downgrade 2 libs
how long is it supposed to sit at Unable to find configuration. Creating default one in ...
Oh ya forgot to mention the tuya device is portable and will surely have internet but will not be in the same lan environment as the home assistant. Therefore looking at setting up the assistant on the cloud or any other way to call an api. Just realised maybe ifttt is a better solution to this
Take several steps back. What do you want to do?
Not, what solution are you trying to implement, but "I want to come home and the lights to come on automatically".
@brazen herald asking me?
You could, and people have
Yup. It sounds like you're going for a really complicated solution right now
can anyone assist with an MQTT issue I'm having? Its hurting my brain 😅
Only if you ask the question
ahh well! better ask it then ha
(This is probably the wrong channel though)
which channel would you recommend?
Depends on whether it's #integrations-archived or #add-ons-archived or even something else
It's pretty unlikely to be to do with installing HA though
no you are correct I shall ask in addons thanks
Hi, my HA on my M1 MacMini runs in general fine but after a week or so I get many errors like this:
And I have to restart hass. Where would I start to solve this?
The starting point is likely to be tracking down the cause of OSError: [Errno 24] Too many open files
What else runs on the Mac? What integrations are you using?
But that could equally well be not the place where it leaks file descriptors. It’s network connections that probably trigger it
No special integrations - just ZigBee
On that Mac there isn’t much going on
I know Pi 1 B+ v1.2 are not supported for haos anymore... I tried installing an older version, i basically just need it to have a single web panel to control some WLED and Shelly relays at one location and happend to get these raspberry pis for free from a friend, assumed that would require near no hardware to just have simple toggles to turn on light and switch colors etc.
Getting this to work seems like a pain tho, the older version of haos just hangs forever with this error if you enter any commands on the prompt:
system is not ready with state: setup
So i tried raspbian and supervised setup, but os-agent doesn't want to install, does anyone know the latest version of os-agent that successfully installs on a pi 1?
Cause i cannot use HACS or something like that in the future and would have to reinstall anyways then
If i need some other integration not available in default homeassistant
Huh?
Yeah
Add-ons aren't integrations
You are aware that there are no add-ons in HACS.
Integrations, and custom components, work on every install method
I throught HACS was an addon, i used to have home assistant running a docker container on my unraid server at home... And had to switch to a VM because i couldn't install HACS
Well, that had nothing to do with the fact that you ran HA in Docker
Think it was the fact that alot of stuff requires AppDaemon or something, which doesn't work in container
only HAOS or supervised
I don't remember exactly
but you're right, docker should be fine then
for this application, just need a web ui to control light stuff for now
AppDaemon works fine outside of HAOS and Supervised
Pretty sure it pre-dates those
It's been a long time tho since i ran it a docker... Think it was back when the "os" was called HassIO or something... I just remember there was something i couldn't do in docker only... So i just swapped to a VM, which has worked wonders since then... Zero issues what so ever. 🙂
Alright, well no point messing with supervisor then if that's the case
Thanks alot
God these old PI 1s are slow tho
even with OC 😛
Is there a way to get the official Home Assistant OS 7.6 running via WiFi. I burnt the SD with Raspberry Pi Imager but it didn't take the WiFi settings.
I know some people think WiFi is a dumb idea bur I have reasons.
The Pi Imager offers the HAOS image but it doesn't work with storing the wifi credentials with that advanced pop up like in RPiOS
is ethernet out of the question at least for the initial setup?
@ everyone is thankfully disabled on the server. If someone would be interested I'm sure they'll reply but don't star tagging everyone with a colour if they don't reply (just a friendly note) 🙂
Sometimes I havn't physical access to the router. And because I'm new to this sometimes I brake my installtion and have to start it from a fresh SD 😦
It would be easier to set up wifi from the UI, that's why I was suggesting it
otherwise you either need to create a special USB with the wifi credentials or add them from the terminal
Ahh, that's possible? I'm just restoring from an 180MB Backup and haven't seen it until now.
In the newest version it's under Network settings
what are you doing though that leads to having to reinstall?
I ran supevised on a RPi4 `You are running an unsupported installation' 😦 and wante to run the VS Code at on. The reason for that is that I have a lot of automations and wanted to put them in a split directory. But it's not possible to edit them via UI there. Hopefully VS Code would support me better.
which pi?
nevermind
blind
You can always just use VSCODE on your computer
and just open your files remotely
and yeah indeed automations not in automation.yaml are not UI configurable
there is one workaround having some split and some configurable
I know. I mounted it as a Samba share. When my installation failed I got trouble with my client. I would prefer to manage automations in directories vi HA UI.
that's good and all until something HA and UI goes down
at least make sure to still have SAMBA, SSH installed and running so can troubleshoot if it comes to that
Though developing in automations.yaml and after testing pushing them to the automation directory?
you basically want to make automations via the UI editor but also have them split up ?
Yes! 🙂
Hmm yeah that's tricky
can't have both
Why not just filter them in the UI when looking for one based on entity/device/area?
If the system is able to access them in directries why not the UI???
Hi all, i have a portable tuya platform button. I want to route the click of this button to call a custom API on the internet. The button will be connected to the internet via a mobile hotspot by the user and will always have internet access. The Tuya platform inherently does not provide an option to call a custom API. Therefore I wanted to setup a home assistant instance to detect the click of the button and route the command to calling the API. To achieve this i wanted the homeassistant instance to also be on the internet (a cloud like AWS server hosting it) so that i don't have to worry about the instance having internet issues at my home and use the more consistent internet of a server. Can some on advise on installation setup on this?
Hi guys!
I was wondering if you could advice me a bit on my installation?
I tried flashing HAOS on my ssd with Balena, but it will not boot from sata, but the same SSD will boot fine with an usb-sata adapter. I have found a couple threads on the forums with people experiencing the same and it seems it being looked into.
I would rather have HAOS instead of running it through VM on Linux.
If i run HAOS with an usb adapter for a while and a new update makes it able to boot on my sata port, could i just switch it over with everything being compatible?
I have sucessfully installed HAOS on Linux with KVM, but i cant figure out how to setup a working network bridge as the manual requires. I have never touched Linux before starting to fiddle with Home Assistant and i honestly don't understand much of it.
Any pointers on this would be greatly appreciated! Thanks!
What machine is the SSD in, that it won't boot from SATA?
It's an older lenovo thinkcentre m92p that has a i7-3770
Bios has support for UEFI and i have disabled secureboot and CSM
I asked this in general first and was directed here. Have anyone successfully mounted a NFS drive to HASSOS? Giving me access to files on my NAS in Home Assistant.
its... not easy and you are going to hate it.... one way is from the debug ssh https://developers.home-assistant.io/docs/operating-system/debugging/
Back for help with HA Blue installation that is unhealthy. Previous help suggested an network issue.
Originally had msg unable to connect to internet to update core/supervisor/OS. I re-flashed which did update, but still have unhealthy system (Reason: Supervisor was not able to update) and I can’t install backup (or update supervisor) as can’t connected to internet.
Blue is cabled into a Deco router for internet. Browser on homeassistant.local:8123 says connection not secure. Not aware of anything else using 443 but can check if someone can explain how. Have a static IP address on router which is configured on HA. Don’t know how to access CLI and can’t install anything as no supervisor/internet.
Where can I start to id issue(s)?
tried to update HA to the latest stable version i did exactly what i did last time now it's asking me to create a account like HA was never installed 🤪
Restore from the backup you took before you upgraded
yeah luckly i have a backup 🙂 but don't know what went wrong in the update proces
docker run -d
--name homeassistant
--restart=unless-stopped
--privileged
-e TZ=Europe/Brussels
-v /docker/Home-Assitant/Config:/config
--network=host
ghcr.io/home-assistant/home-assistant:stable
ran this and worked fine last time
You should be using a compose file
Does /docker/Home-Assitant/Config exist, and did you typo originally?
i know but stil havent figgerd that part out how to make a compose file 😅
i does it even changed files in the config folder
Using a compose file means when you screw up spelling you don't "lose" your config
So, it's not actually Home-Assistant/ but mis-spelt Home-Assitant/?
it's Home-Assistant on my nas
Well, there's your problem
You can't spell
-v /docker/Home-Assitant/Config:/config \
^^^^^^^^
no i suffer from dyslexia most of the time i double check this one got away from me
And that's another reason for using a compose file
Files tool away for another day
'tis a handy site for helping people move from fragile command lines to robust files 
so what is my best course of action now ? Restore backup and then use compose file to get things started ?
Check your path
You've claimed both of these are valid:
Home-Assistant/
Home-Assitant/
One of them isn't right - you have to work out which
Home-Assistant the first one is correct
You don’t need to really restore from backup, you can switch to compose right now and keep using the same volume you defined with -v
Yup, but the very quick fix is to simply fix the typo
Hell briefly glossed over the docker cli and went nope and jumped directly into compose
docker: Error response from daemon: Bind mount failed: '/docker/Home-Assistant/Config' does not exists.
oh man i'm no good in this ....
I feel I’ll use it when I feel lazy and a service gives a docker cli as the only example lol
how do i check the direct path if it exists
ls -al /my/path
You can navigate there like normally in Linux too (using cd)
And here I have been translating it arg by arg into a compose file
Why on earth. HA even gives an example config
my path does not exist ....
Or did you mean other things @raw hedge
Well not for HA… but I still did it from compose to k8s https://github.com/miwagner1/k8s-at-home-with-k3s/blob/main/stacks/homeassistant/base/deployment.yaml
I translated the example config
Yeah that’s fair
found it 🙂
its /volume1/docker/Home-Assistant/Config
it"s up and running again
man your a life saver
So… for ease of use and possibly lazy backups, I would look into making a docker directory inside your home folder, make a homeassistant folder inside that and then make a data folder for all of HA’s data
Throw a docker compose yaml in homeassistant
so now i take the code i used in to composerize.com
Then you can lazy samba grab the directory every time you want to make a backup or zip it up, or use a backup solution
I am guilty of lazy samba pulling
You can chuck it here if you want us to check
rsync
rsnapshot FTW
i used composerize :
version: '3.3'
services:
home-assistant:
container_name: homeassistant
restart: unless-stopped
privileged: false
environment:
- TZ=Europe/Brussels
volumes:
- '/volume1/docker/Home-Assistant/Config:/config'
network_mode: host
image: 'ghcr.io/home-assistant/home-assistant:stable'
have me this
I use git now ftw
Ooh I gotta look into that
Uses rsync and gives you a history
Might have to migrate from pure rsync to that
so now i created the file docker-compose.yml where do i upload it
A shame I have to use NFS or a S3 (selfhosted) for my backups now
You don’t upload anything. You stop your existing container, and run ‘docker-compose up -d’
If you don’t have docker-compose you may have to install it
Anywhere, I would make a ~/docker/homeassistant folder and dump it in there
Or that yeah ^
Or just ~/docker
can i dump it in the existing folder ?
Heck, stick it at the root of /volume1/docker/
i use /docker/compose/
My data structure for my home automation containers is all /app/xyz with a docker compose in the root /app directory
okay
Yah but now you have to tab 3 times if you have each stack in its own folder
i have one compose file that includes all my services
and how do i run it next time i need it ?
Then again, I’m one to talk with my new config and overlays
What I said, docker-compose up -d
/docker/containers/<container_name> for all my persisted files
V2 is docker compose, v1 is docker-compose
so only docker compose up -d
Try it yes
no file path
If you’re in the directory of the docker compose file
Oh you need to be in the directory, you can always define a file path for the command
and i change directery with ls ?
thats just listing
or i'm a wrong
cd like in dos
Linux command line is similar but different. So shouldn’t be hard to learn
docker-compose -f path_to_docker_file_and_compose/docker-compose.yml up -d
2 options for you
Luckily I won’t give you 10 like windows does
ran the command it gives a error but that's okay
Windows is nice like that though. Since it means you don’t have to adapt to dos/powershell commands
What’s the error
What?
ERROR: for homeassistant Cannot create container for servi
nt" is already in use by container "138116720e2c028d654b346
e (or rename) that container to be able to reuse that name.
yeah thats why i said it was okay 🙂
Awesome!
so i just shut it down and delete it directly and then run the compose file ?
Well @raw hedge @humble mirage we upgraded someone to use compose tonight 😂 another one to the dark side
Just docker container stop [comtainer name]
Then start it
Should be happy with it
so no more deleting the container sweet
First it’s compose, then it’s k8s
Yah, compose will manage it, make changes to the compose file? Run the command again
With a compose file updating HA becomes just:
docker-compose pull
docker-compose up -d
And there’s a purge command isn’t there to get rid of old images?
I have a script automating it so I can just run Dockerupdate and it updates all my containers so I forget lol
Yes, or you can use another docker deployment to auto purge old unused images
K8s auto purges unused images 😂
It’s a minimal 550 Mb of ram overhead for masters that are also workers by the way
Make sure you have enough ram :p
The only dark side is Supervised
I am running Debian virtualised on a machine with 32gb ram, think I can spare 550mb 😛
Cloud hosted supervised?
question non related but i can then also make a compose file to update my minecraft server in docker ?
Start with k3s
Yes
One compose file, all the containers
If the server is docker then yes. Can put it all in the same file
Then if you do docker compose pull it’ll update all of the containers in the file
I tend to like to put each stack as it’s own compose file
And below services :p
So services is the overarching part, each under part is the service
Any New York or New Jersey enthusiasts here?
yeah so if i need to add a new one i just do
services: bla bla bla
You are using docker compose version 3.3 and you are defining a homeassistent service
So next you want to add a Minecraft service
So you start with the service name
It's a YAML file, all the same rules as for HA's YAML applies
Add the rest
Here is a cool thing with docker, inside that compose yaml, the services can talk to each over using their service name
So you can have a mqtt service named mqtt and ha can talk to it with the dns name mqtt
No need to use your ip
sweet
Take a look at atx’s config to see how a full docker compose can be https://github.com/atxbyea/docker-compose
now i just need command to update my minecraft and i'm set
Can give you ideas of how it works
Oh I need to show you how full it really can be
Oh yeah it’s just an example that I knew existed
I used to instead of using bind mounts, I used named volumes but I define the bind mount in the named volume properties
Then I could add samba container and just used the volume name to mount the data
You can also mount NFS shares etc this way
@manic bane see https://imgur.com/a/J6vzr2G
This was when I was trying to use glusterfs only to find out that using glusterfs and docker together caused FUSE to crash which un mount the volume… and this is a years old issue
My drive for k8s was to use longhorn for storage :p
That was what made me move away from docker compose / swarm
The only thing thriving left for docker is their hub, docker desktop for dev work like building images, their registry and docker compose
Which is good they are focusing on that since they cover k8s devs, and non cluster devs/users
i assume i have to write them all in the same version in the file
You don’t keep adding a version
You already defined that, you just need to define the service name or aka just copy everything under “services”
ill try my best
Use the tool, pastbin it and share here and I can take a look
services:
minecraft-server:
ports:
- '25565:25565'
environment:
- EULA=TRUE
image: itzg/minecraft-server
it does not stat volume thats the code i got to copy and past
so it's not correct i think
can you link persistent storage without full path ?
instead of /volume1/docker/Home-Assistant/Config:/config
you use ./Home-Assistant/Config:/config ?
or is it always bether to use full path
yes ./Home-Assistant/Config:/config will work

