#installation-archived

1 messages Β· Page 71 of 1

humble mirage
#

Try HAOS 12.2 or 12.3

weak sand
#

how do I do that ?

humble mirage
#

Download the 12.2 or 12.3 image, from the relevant release page

weak sand
#

thanks ! πŸ™Œ

#

Sorry I'm not used to GitHub, again a new thing for me. Where do I need to go ? I went in Home Assistant OS 12.3 https://github.com/home-assistant/operating-system/releases/tag/12.3 but then I don't know what to do. In Bump buildroot to update linux-pam (#3325) ? If yes, I don't know what I need to type nor download... Sorry, it's some idiots questions x)

humble mirage
#

Scroll down to Assets

#

Find the HAOS image that matches the one you already downloaded

#

haos_ova-12.3.vdi.zip for example

weak sand
#

Yes I downloaded this one, thanks, And then I create a new VM ?

humble mirage
#

That or change the current VM to use it

weak sand
#

okay boss, thanks I'll try that

brazen harbor
#

You were right by the way, thanks

frozen citrus
#

😒 Cannot find or remember the password for a new install, I have my ha running on a usb ssd.

Any pointers as to the best way to recover/change without losing my configuration?

#

the web password

raven forge
frozen citrus
humble mirage
#

You're still fine πŸ˜‰

#

Maybe take this as a hint that you should use a password manager

#

You can see the users in /config/.storage/auth_provider.homeassistant

frozen citrus
#

Thanks for that, I do have a password manager however I have 2 systems running in different homes and the password manager saved the credentials for http://homeassistant.local, I have the credentials of the other system 😒

frozen citrus
#

i do not know how or if i can access the file from the haos as it does not seem to have directory commands

raven forge
#

Run auth list in the CLI like described in my link.

frozen citrus
#

Thanks all resolved πŸ‘πŸ‘πŸ‘

subtle orchid
#

i need to add a custom dns to home assistant to connect it to the internet

#

how do i do that ?

proven grove
subtle orchid
#

it can't install it just throws errors because it can't connect to websites

proven grove
#

Ohhhhh... hmmm πŸ€” Not sure, tbh. Is this on a rpi or a VM or something like that?

subtle orchid
proven grove
#

Right, yeah. Any reason you can't just attach it directly to your router and let DHCP take care of it?

subtle orchid
#

actually

#

it sometimes works

#

but i still need dns assigning

proven grove
#

Is the rPi getting an IP address?

subtle orchid
exotic hare
#

Anyone have luck installing HAOS on AtomMan X7 Ti because I cannot get it to boot? I use Try Ubuntu on a USB to load the HA image on my hard drive. I cannot figure out why it will not boot up. Also, I ran the efibootmgr command as well.

raven forge
#

What happens during boot?

exotic hare
raven forge
#

Can you boot into the ubuntu ISO again and run and share the output of these commands?

efibootmgr
lsblk -o+FSTYPE,LABEL,MODEL

You might have to prefix them with sudo . I don't think it's related to it but did you disable Secure Boot? Have you tried slot B?
With a x86 machine it can make sense to virtualize. Have you thought about doing that?

feral lichen
#

Hi, I'm completely new to HA, I have installed HA using HA-yellow and able to configure. Due to moving to a different location I have removed all the devices that I have added earlier. After moving to new location I'm not able to launch http://homeassistant.local:8123/ url and when I try to do getting an error this site can't be reached. How to resolve this issue?

raven forge
#

Try the ip instead.

desert heath
#

Hey I tried to install the Home assistant operating system on my x86 hardware, about 18 months ago. But my hardware was too new to be supported. I feel like its probably a good time to go back to it, I dont want to have to deal with maintaining the computer any more. The install instructions say to use ubuntu to boot the system and write the image to the disk, but this caught me out last time, because ubuntu did have support for the hardware. Is the homeassistant os built on debian though, so I can use a version of its live cd instead? if that wont boot I wont be able to get to installing it.

#

how do I figure out what version of debian underlies the homeassisnt os image, to test its support on my hardware?

#

lol, the image is compressed with xz!

raven forge
#

HAOS is not based on debian.

desert heath
#

ok, so its its own os, so it very likely might not have support for my hardware.

raven forge
#

What hardware do you have?

desert heath
#

some generic minipc thing.

raven forge
#

You can always install proxmox VE which is based on debian and virtualize HAOS. I recommend that anyway for anyone who is brave enough.

desert heath
#

I dont really want to use proxmox

#

but I can use debian and supervisord?

raven forge
#

You can install debian and run the docker container version.

#

Supervised is not recommended.

desert heath
#

why? I have to maintain the os myself?

raven forge
#

Yes. In a very specific way.

desert heath
#

I would need to do that with proxmox or the docker container already. I have been running the supervisord version already on ubuntu (I know thats unsupported). My goal here is to get back that part of my life where I have been maintaining my private port.

raven forge
desert heath
#

this box used to do other things as well, which I know is unsupported. I have moved that all away now.

#

I was busy setting up my house and new hardware takes weeks/months to arrive here.

#

thanks for the advice, I came to get it, let me think whats best!

#

I think, I am going to take it. my issue with proxmox is the machines ram

#

My hope is I can get hasos running on it this time around.

raven forge
#

PVE needs about 1.2G or so for itself.

desert heath
#

yeah, its got 4Gi

#

poking around the image now, to look at what the kernel really is.

raven forge
#

HAOS should work on any EFI supporting x86 machine if you disable secure boot.

raven forge
desert heath
#

I was supprised when it didnt run, Its a celeron cpu, so I dont think thats a problem, realtek network, is older than the sun.

#

I am gonna try haos again, if It doesnt run, I will put back the backup I make of the disk and figure out what to do next

blazing folio
#

Hello, looking for some help debugging my home assistant which is not starting (running on a raspberry pi). I've logged into the OS with ssh root@homeassistant.local -p 22222 and it looks like the containerd service fails to start. Has anyone experienced this issue before? See journalctl logs in thread

urban tusk
#

Hey all, I'm experiencing (apparent) crashes of HAOS. Symptoms are the UI just doesn't load any more or produces a lot of 500s. How do I go about debugging that?

humble mirage
#

What hardware is it running on?

urban tusk
#

I'm actually pretty fluent in linux, but the way HAOS is set up, it doesn't really give me a lot of options to get in there and sort things out.

#

It's currently in crashed state, so if anyone here wants any volatile info before I hard-reset the device, please speak up.

#

I'll wait until 10:00 utc, then I'll reboot it.

urban tusk
#

any way to get logs from before the reboot?

#

Never mind, found them

#

Nothing in the logs at all. It just seems to suddenly die at 9:33 this morning with no warning at all. Next entry is the reboot.

#

Any ideas what I could do to debug this issue?

#

Anyone?

dawn dome
#

hey guys, I started my ha journey yesterday and have stumbled upon some problems which I'm not able to fix. To start off immediately after installation I got some errors regarding the default_config and some dhcp problems. I've also had some problems regarding mqtt but feels like it traces back to this. Is there anyone who would be able to help me or guide me in the right direction. Thanks a lot

The following integrations and platforms could not be set up:

dhcp (Show logs)
default_config (Show logs)
Please check your config and logs.```
#

I am running haOS on a rpi 4 with an external ssd

proven grove
humble scaffoldBOT
#

@dawn dome Please use a code share site to share code or logs, for example:

Please don't use Pastebin, since it can randomly add spaces to the main view. Please also don't share text as images since it makes it harder for people to help you. Remember that others may have colour blindness, impaired vision, etc.

dawn dome
#

it's almost empty, don't know if it is meant to be this way

#

Based on one of the errors it seems like scapy.config might be a problem but I'm not too certain on how to fix thisSetup failed for 'dhcp': Unable to import component: cannot import name 'conf' from 'scapy.config' (/usr/local/lib/python3.12/site-packages/scapy/config.py)

proven grove
#

Yeah, that's odd. This is a new install, so I'd recommend flashing again and reinstalling. Something didn't go right the first time.

dawn dome
#

I'll try that, thanks

glossy palm
#

Does mDNS discovery in HA specifically not work in the docker container? If I install avahi-tools in the container for testing and use avahi-resolve host.local I can get an IP but ping host.local fails and I cant use mDNS names for things in HA

proven grove
glossy palm
#

I am not using docker and I cant see any mDNS queries coming out of the container at all using tcpdump, can see them coming from esphome and other things deployed the same way though

#

I mean using avahi tools I know mDNS works...

#

I know why, its a musl problem

#

I forgot HA was built on alpine

#

I am guessing there is no official non-alpine container?

humble mirage
#

No

proven grove
#

Nope.

humble mirage
#

Also, it's probably more that it's not set up to use mDNS for general resolution, which isn't unusual

glossy palm
#

Nah, musl lacks the equivalent of libnss-mdns so the system cant properly resolve mDNS addresses properly in a lot of cases, there is a work around which uses avahi and a mini dns server from memory

#

Docker network mode host basically uses the host OSes resolution which is likely not a musl based distro and hence works

urban tusk
#

Hey guys, I really need help here, so throw me a bone. Why would HA just stop responding? The http server seems to be up after a fashion but serves nothing but the HA logo. There's tons of crap in the logs, but nothing spelling impending doom. How can I debug this issue?

proven grove
humble scaffoldBOT
#

Please use a code share site to share code or logs, for example:

Please don't use Pastebin, since it can randomly add spaces to the main view. Please also don't share text as images since it makes it harder for people to help you. Remember that others may have colour blindness, impaired vision, etc.

urban tusk
#

Thing is, it's the same crap all over again. Some division by zero, some deprecated apis in some integrations, that kind of stuff

humble mirage
#

Share it

urban tusk
#

I can just give you the log file right here, can't I?

proven grove
humble mirage
#

FFS

humble mirage
#

You've only been here a month, surely you can't have forgotten already πŸ˜›

urban tusk
#

It's a PITA if you have a file instead of text...

humble mirage
#

Help us help you... downloads of random files are something few people will risk

urban tusk
humble mirage
#

Have you tried starting in safe mode, to disable all those custom components?

urban tusk
#

Not yet. Problem is, it only happens every few days or so.

#

If only I could get at the bloody journalctl...

proven grove
#

You're using some kinda crappy custom integrations, tbh and that might be part of the problem here. MIGHT

urban tusk
#

Ok, I'll run it in safe mode for a few days, but the problem here is, as rare an occurence as that problem is, that won't actually prove anything besides how much my setup depends on those custom integrations.

#

That being said, I could PROBABLY get rid of all the custom stuff EXCEPT AlexaMediaPlayer and, at least until I figure out how to get the replacement running, SureHA.

proven grove
humble mirage
#

AMP is on life support/dead

urban tusk
#

I am painfully aware of that, yes.

humble mirage
#

Basically the goal here is to find which of those are causing you problems and report the issues in the hope they get fixed

#

Or move to other solutions when it becomes clear they won't get fixed

proven grove
#

(AMP probably won't be fixed any time soon)

urban tusk
#

Frankly, there's nothing to fix in AMP, unless Amazon does a 360. I'm looking into getting that replaced, but that's going to mean finding a replacement for all my echos.

humble mirage
urban tusk
#

Haven't really had the time to look into it yet.

proven grove
urban tusk
#

Only a few weeks ago Amazon just pulled the rug from under all the list integrations.

humble mirage
#

The problem here isn't the integration with Amazon, it's with the code quality

urban tusk
#

Well, I can't judge that, I'm not a Python guy at all and never will be. Even me writing a yaml file looks like hand-to-hand combat in a chicken coop.

#

I know many people love it, but I never quite got my head around the idea of structuring code based on indentations.

#

Give me a good C-oid language any day...

languid goblet
#

I don’t have perms to upload images in this channel, I’ll just link them

humble scaffoldBOT
#

@languid goblet 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.

languid goblet
#

2024-07-25 05:34:09.412 ERROR (MainThread) [homeassistant.setup] Unable to set up dependencies of 'default_config'. Setup failed for dependencies: ['zeroconf']
2024-07-25 05:34:09.413 ERROR (MainThread) [homeassistant.setup] Setup failed for 'default_config': Could not setup dependencies: zeroconf

proven grove
#

And this is a non-official docker image, right?

languid goblet
#

Correct, non-official, it’s from TrueCharts

#

I think I’ll just reinstall TrueNAS and do the ram and drive swap, leaving me to spin up Home Assistant OS via a VM

#

Internet assures me that I shouldn’t lose data from a ZFS data pool export

proven grove
#

Yeah, that sounds a lot better.

#

I can't 100% confirm that (been a LONG time since I've run TrueNAS).

languid goblet
#

I needed to do it eventually as there was no Matter docker integration like in Unraid (it worked fine there)

#

As nice Unraid was, it’s a $60/year license, no thanks lol

#

I’m trying to move away from subscriptions, not gain them

#

Thanks for your help!

proven grove
#

$60/year? I pay $36/year for mine.

urban tusk
#

Well, I managed to get rid of SureHA.

desert heath
#

heya, preparing for a re-install, if I take a full backup a restore from it is all adons also? or just HA itself?

tender wren
#

Incl. all add-ons (if you download it before reinstalling HAOS)

desert heath
#

I just ran it, and downloaded it to my computer, but I think its stored in my google drive as well.

humble scaffoldBOT
#

@exotic hare I converted your message into a file since it's above 15 lines :+1:

πŸ“Ž message.txt
maiden mauve
#

I"m sure many here are proxmox gurus so I'm hoping somone can help
I recently updated from proxmox 7-8 and started experiencing random shutdown of the HA docker

I'm not experienced enough to pull up the logs to see what could be wrong.
HA logs aren't showing anything useful

humble mirage
#

How are you running HA? LXC? VM? Other?

#

Bets on ||tteck's scripts||

raven forge
urban tusk
#

@proven grove how's YOUR Python-fu? I mean, I'm a decades-long developer and have worked with AWS a lot, so I know the Amazon style... If SOMEONE gave me a bit of a leg-up and an open ear for my Python-nooby problems, I might be able to do something with AMP

proven grove
urban tusk
#

Well, refactoring doesn't sound TOO bad, with a bit of a leg up.

#

I've refactored Fortran code at one time, although I admit that was one of the most scary experiences of my life and I had NO idea what I was doing...

proven grove
#

Eh, pull the latest and start exploring. Pay attention to the async_* calls and then reference what they should be in the dev docs.

urban tusk
#

I'll have a squizz, but don't expect me to do anything meaningful. SOMEONE is doing something to it, PRs are still being processed.

#

About a dozen last weeks alone.

proven grove
#

Yeah, there's been some people trying to get it into a useable state.

urban tusk
#

Yes. the code owner to start with.

#

Code frequency is the highest since january,

#

Looks like he's getting back on the old horsey.

#

@proven grove just to be clear on this: the problems are definitely in the module itself, not hidden somewhere in alexapy?

proven grove
#

There might be issues in alexapy. But the ones that really affect HA are all in the threading mostly. When it was written, HA didn't really care so much about it and let integrations pretty much do what they wanted. It's much more structured now and a lot of async calls have changed and/or been marked deprecated.

deep basin
#

Hello, would someone be able to confirm for me if my current Hoobs Pi board will be able to run HA? I have a BPI-M2-BERRY V1.0 From what I can see these are the specs

Banana Pi BPI-M2 Berry is a quad-core mini single board computer built with Allwinner A40i SoC. It features 1GB of RAM . It also has onboard WiFi and BT. On the ports side, the BPI-M2 Ultra has 4 USB A 2.0 ports, 1 USB OTG port, 1 HDMI port, 1 audio jack, and last but not least, a SATA port

I got the specs from https://wiki.banana-pi.org/Banana_Pi_BPI-M2_Berry

humble mirage
#

Home Assistant, yes. HAOS, no.

deep basin
#

oh, 2 different things huh, ok. Lemme go look what the differences are. TY

humble scaffoldBOT
humble mirage
raven forge
humble mirage
#

Nope

ebon ice
#

Oh, could be πŸ˜„ Is it over 15 lines πŸ˜„

humble mirage
#

21 >15 πŸ˜›

ebon ice
#

Oh! the empty ones are lines too... makes sense πŸ˜„

raven forge
humble mirage
#

Roll on the server revamp

raven forge
#

Tellmehow.gif

ebon ice
urban tusk
raven forge
#

I'm not sure.

urban tusk
#

@humble mirage what do you think how would people feel about changing that line count enforcer up a bit? Specifically to make it preserve more of the message structure. If I have a message like

"Here's my code:

more than 15 lines
```"
to not include the header line in the file, but put it in the message directly?
#

Frankly it's annoying to have code and non-code just rolled into one ball.

proven grove
#

Honestly, no. Would rather see it convert the message over to dpaste.org. (Just my opinion).

urban tusk
#

Hm, might look into that, but that would take a lot longer.

humble mirage
urban tusk
#

The logic I would propose is this: Count every code block as one line. If that still is more than 15 lines, roll it all into one. If that is less than 15 lines, just roll the code blocks into files and leave the rest as is.

humble mirage
#

But given that's hard the alternative is a re-work of the server to move to forums, so that walls of hate text don't mess things up for everybody

#

But ... people will still fail to use code markup, so it'll all still be an unreadable mess

#

Also, the bot's options are all or nothing, the bot can't edit posts

raven forge
#

I'd prefer if it rejects/deletes the message with a link to those rules so people keep the agency over their messages' content. As in being able to edit it.

urban tusk
#

And the way I read the code, what the bot actually does is REPLACE the message. And since it has full control over the message content, it can do what I propose. Just Git-style: don't edit, replace.

humble scaffoldBOT
#

@urban tusk When using Discord's Reply feature it defaults to pinging the person you reply to, which can get frustrating for the target. Use Shift + click on the Reply option, or 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).

urban tusk
#

Anything else I can do to avoid annoying you?

humble mirage
#

Sacrifice an idiot at the alter of wisdom?

ebon ice
humble mirage
#

That's not "expected", but it does seem to happen a fair bit

ebon ice
#

huh πŸ˜„ well, hmm, huh. Is there a way to handle it? I mean, I have internet connection πŸ˜„ not sure what the issue is πŸ˜„

humble mirage
#

It's probably DNS related, but hopefully somebody who understands HAOS will be able to help you fix it

urban tusk
ebon ice
#

changing DNS did not work. Running this command ha jobs options --ignore-conditions internet_host did not help too. God damn πŸ˜„ I mean, seriously πŸ˜„ πŸ˜„ πŸ˜„

urban tusk
#

Or as my driving teacher put it: "If I had a rocket launcher in the car, I wouldn't have time to drive at all with all the reloading."

humble mirage
#

See... you just need a tank with an auto-loader

ebon ice
# 
➜  ~ curl https://version.home-assistant.io/online.txt
curl: (6) Could not resolve host: version.home-assistant.io
urban tusk
#

I think you are on to something. If we combined the autoloader with the automated reprovisioning system from the PzH 2000, and put the resupply apparatus on a trailer pulled behind the tank...

ebon ice
#

when I type this command ha dns info
I'm getting this:

fallback: true
host: 172.30.32.3
llmnr: true
locals:
- dns://8.8.8.8
mdns: true
servers: []
update_available: false
version: 2024.04.0
version_latest: 2024.04.0
#

That is not my hosts IP?!

#

looks like a docker ip

proven grove
#

That is a docker IP. In HAOS, Core is a container.

raven forge
#

Check ha dns logs. Perhaps disable fallback. Also try the common stuff like ha network info and ping 1.1.1.1.

ebon ice
#

found out when network IPV4 is set to automatic, it works of course. The second I add static, and add the exact same settings like I'm getting with ha dns info, it stops to work.

#

While automatic, the command ha network info gives this host_internet: true, but when Static IP i get host_internet: false

raven forge
#

Please share the full output of both versions.

humble scaffoldBOT
ebon ice
raven forge
#

Hmm. Can't see anything obviously wrong. I noticed that both have supervisor_internet: false.

#

What does ping 1.1.1.1 say with the static ip?

ebon ice
raven forge
#

I assume it doesn't do this with the DHCP ip?

#

Is it possible the ip you gave it is used by something else already? The static ip was different.

ebon ice
#

The IP is locked for my HAOS πŸ˜•

I'm pinging other local devices without issues

#

dunno, should I reboot my router too?

raven forge
#

I never really had to do that but it probably can't hurt.

proven grove
raven forge
#

I'd recommend that too but I'm very curious about why this happens.

proven grove
#

Fair... so am I actually.

ebon ice
#

would you look at that... the router reboot did it...

#

I'm not using DHCP reservations as I had bad experiences in the past with my router. It did not worked always as it should etc 😦

proven grove
#

Sounds like you had a bad router. 😦

ebon ice
#

TP-Link ftw....

#

ok, but still getting this HAHAHAHA omg... 'BackupManager.do_restore_full' blocked from execution, no host internet connection

#

Ok, as far as I see, this command helped now:
ha jobs options --ignore-conditions internet_host

#

Looks like the migration wokrs now too after the internet was established... ffs πŸ˜„ ❀️

gloomy orbit
#

Hello, maybe this is leaning towards a proxmox questions more than a HA one, but could anyone hint at what is going on here? Glances show memory usage of 1.9 gb, while proxmox is indicating over 7gb
https://imgur.com/a/YeCpv8H

#

Hmm, it seems the buffer/cache is using almost all of it. Maybe Glances report without buffer/cache use? Is this something to be worried about?

From running top in the terminal:
https://i.imgur.com/LNK5FaU.png

raven forge
gloomy orbit
raven forge
#

Kind off. I'd only alllocate as much as needed.

#

PVE will remove memory (to min) from the VM if the host reaches 80% and you use ballooning. Inside the VM cache is pretty much instantly available when needed.

gloomy orbit
#

I think I understood that... πŸ™‚ thanks for clearing it up!

gloomy orbit
raven forge
#

There's also a graph in the system settings.

urban tusk
#

@proven grove re what we talked about earlier, turns out dpaste.org does have a sort-of api. leveraging it in TS would be somehwat messy, but doable.

raven forge
gloomy orbit
raven forge
#

Yes but you'll probably swap at some point. Swapping is fine but swapping because of lack of memory is not. I'd probably give it 6G or so (5G just feels wrong). Perhaps 4/6 min/max. See glances to check what uses "so much" memory.
No need to be sorry, btw.

gloomy orbit
#

Great, thanks. I'll probably need to read up on how memory works. It's been a while. πŸ˜„ Right now I'm on 1.9g usage which I suppose is pretty normal. The HA container itself uses 1.1, otherwise its grafana and frigate using most of the rest.

raven forge
#

Yeah it's pretty install/environment specific. I give my HAOS test VM 1/2G but that's not really running much. My docker install uses about 900M.

old night
#

Hi All, I have a question regarding installing HAOS on RPI 4B , I successfully installed it , but I'm struggling on how to install the RPI touchscreen as each time I try to use ssh to install it either I face apt not found or if do it manually the system remove it when restart, can you kindly guide me on how to do that?

proven grove
old night
old night
proven grove
urban plover
#

Hello everyone! I'm attempting to install HAOS as VM on my ubuntu server. I'm using KVM and virt-install. I can access http://IP:4357 but not http://IP:8123. Nothing meaningful is showing up in the logs. Looking around here & google this seems to be a rather common problem but I haven't found an actual solution yet

copper steppe
#

Is your network bridging to the LAN, or NAT through the ubuntu IP ?

urban plover
#

I have it NAT. But I am curling from itself, the HAOS vm. (and port forwarding 8123 currently, but can't even access the webpage internally)

copper steppe
#

check the console logs in haos and what it thinks. maybe the port is already used?
try bridge, then it acts more like a real machine on the LAN, rather than some service on the box

#

what erro? con refused, or ssl err, or something else?

urban plover
#

I'm going to try a reinstall and look at the logs again before I go to bridge. I have some other networking stuff going on my ubuntu server and I'd rather not connect it with a bridge. I see some weird http error in logs but then I restarted core and never got it again.

copper steppe
#

paste "network info" output from haos cli

urban plover
#

Curl for :8123 returns empty

humble scaffoldBOT
#

@urban plover I converted your message into a file since it's above 15 lines :+1:

πŸ“Ž message.txt
copper steppe
urban plover
copper steppe
#

Curl -v

urban plover
copper steppe
#

nah, only 1-2. but, for me, I want a VM to look on the LAN like a real machine, not a service on the host. so I can move it around, and it'll still have the same IP etc.

vernal wind
#

hihi so i am helping set up a smart home and was directed towards Home Assistant. I dont have another system to dedicate to HA at the moment. Im usuing virtual box to run it but on installation i recieve the "jump to emergency console" error message. Any help would be appreciated

vernal wind
copper steppe
#

You'll need to provide more information than that. Are you failing during install, or subsequent boot? what messages are seen on console? it's 90% a networking issue.

#

have you done the web stuff during install? you need to login etc

vernal wind
#

after some research i think i just have to have my Hardware build connected to ethernet even if on a VM

lunar siren
#

I just got a HA green. Plugged it in and am trying to go to homeassistant.local.8123 and Im getting nothing.

copper steppe
#

it's :8123 not .8123

lunar siren
#

oops thank you but unfortunately same

copper steppe
#

I don't know anything about green specifically, but iirc its just a sbc, right. Does it have a console you can check? Can you ping homeassistant.local? Can you check your router to see if the green got an IP address?

lunar siren
#

green doesnt have an IP in my router but i can ping homeassistant.local

copper steppe
#

uh, those two answers don't make sense. it must have an ip address to be working at all. I don't mean a static ip just any ip.

#

.local is mdns/zeroconf/avahi stuff to help you find the address

lunar siren
#

over my head probably

copper steppe
#

if you ping it, what ip address does it say?

#

Somethign on your lan? or some stange private-network thing?

#

mine is, for example,

❯ ping homeassistant.local
PING homeassistant.local (192.168.30.33): 56 data bytes
64 bytes from 192.168.30.33: icmp_seq=0 ttl=64 time=0.917 ms
64 bytes from 192.168.30.33: icmp_seq=1 ttl=64 time=0.977 ms
lunar siren
#

it doesnt give me an IP like that

#

Pinging homeassistant.local [fe80::eb06:4f8:89bd:446a%7] with 32 bytes of data:
Reply from fe80::eb06:4f8:89bd:446a%7: time<1ms
Reply from fe80::eb06:4f8:89bd:446a%7: time<1ms
Reply from fe80::eb06:4f8:89bd:446a%7: time=1ms
Reply from fe80::eb06:4f8:89bd:446a%7: time=1ms

copper steppe
#

that's ipv6

raven forge
#

Try ping -4 homeassistant.local then.

lunar siren
#

ahh ok that gives me 192.168.1.255

#

but then times out

copper steppe
#

255 is not a real addres, that's a netmask. I don't think it's valid...

lunar siren
#

is there a way i can just start over?

copper steppe
#

I don't know green, but you could probably configure your router to give the green box a static (and different) ip address

raven forge
lunar siren
#

ok thank you, ill try that

#

ok im hooked up and it is in fact showing 192.168.1.255

#

static

cinder marsh
#

.255 is broadcast ip

copper steppe
#

yes, it's not a valid host IP

cinder marsh
#

It shouldn't be that ip basically. You should change it

#

Try those instructions but change the IP address to one that isn't in use ie. 192.168.1.250

copper steppe
#

anything that's not currently used, and is not .0 or .255

lunar siren
#

having trouble. directions dont work for me. it cant find the interface

copper steppe
#

you have it plugged into ethernet?

lunar siren
#

yes

copper steppe
lunar siren
#

nope, i got that up

#

im trying to do the network update part

copper steppe
#

but the ipv4 one says .255 in your system information?

lunar siren
#

correct

copper steppe
#

is the led on the ethernet port green?

#

I don't know why it would get 255 unless your router is giving it crazy address

lunar siren
#

trying something else i found

#

yea its running and all

#

ok i think maybe i got it

#

i got the IP to be blank now

#

and it doesnt look like i can factory reset because i cant get into the portal 😦

raven forge
#

ha os datadisk wipe should reset most things.

lunar siren
#

ok let me give that a go

#

well i think something bad happened. its just sitting there at a blinking cursor now

weak sand
#

Does someone know how to install HA with a VM but without virtualbox and Hyper-V (because i have windows 11 so can't Hyper-V) ? Maybe one that can clone my VM to put it on another computer like VirtualBox does it

humble mirage
#

VMWare?

weak sand
#

Is it free ?

humble mirage
#

No idea, it used to be

weak sand
#

ok i'll look into it thanks !

humble mirage
#

Why is VBox not an option?

weak sand
#

I'm the one trying to put HA on an administrator computer wihout internet. For some reason, I can't add a zigbee dongle, HA detects it well but when I click to configure it, it says 'loading next step for ZHA' and never stops which crashes my VM. So I bought a new dongle since VirtualBox and HA were working before that, but it is still not working. So the only things I can change are VirtualBox and HA so I'm trying to change VirtualBox since I have more hope in HA...

humble mirage
#

Maybe you should stop using Windows πŸ˜‰

weak sand
#

Since it's a computer from my company I can't really put linux (don't really know if I can't and how to do it...) but maybe I can create a linux VM on virtualbox andthen downoald in it virtualbox and create HA VM. It's a bit strange but you think it can work ?

humble mirage
#

That seems likely to just cause you more problems

weak sand
#

ah ok ... I'll try VMware before asking my boss for linux πŸ™‚

#

Oh I had another question, When I try to power up my HA VM with virtualBox or Hyper-V, it goes in emergency mode because HA CLI not starting. Before that, I thought it was because the computer needed internet, but now it has internet and it still does it. That's why I'm used to create my VM another computer and then clone it... Here is the logs : https://imgur.com/a/it2Nyqn

humble mirage
#

Pictures of text are horrible

weak sand
humble mirage
#

It's still a picture of text...

#

Text shared as text is so much easier to deal with

olive ember
#

may be you could recycle an old HDD drive, and boot on it?

weak sand
#

I don't have that...

weak sand
olive ember
#

Your errors clearly says that google dns cannot be reached. you have the ip 172.30.232.2 looks like virtual box is creating a new virtual network for your VM, but this network is not NATed, or is it your real lan network ?

weak sand
#

But I clicked on bridge network it's strange, this IP is not the IP i want to bridge it to

olive ember
#

may be u updated virtualbox and some default behavour have changed

weak sand
#

I just downloaded virtualbox and instant create my HA VM, same with Hyper-V. And I have this error only with this administrator computer. It doesn't have much things in it so maybe it creates a problem but thanks anyway !

raven forge
olive ember
#

.... ..... .... "Bridged networking. This is for more advanced networking needs, such as network simulations and running servers in a guest. When enabled, Oracle VM VirtualBox connects to one of your installed network cards and exchanges network packets directly, circumventing your host operating system's network stack."

Sounds good, but not working ? A quick reboot ? U selected the right network card ? is your VM using a static ip ?.....

raven forge
#

Check ha network info.

weak sand
#

Since I go in emergency mode, I can't write network info... Yes I selected the right network card. I don't know if my VM is using a static ip since I cannot see anything but I think it's in automatic mode...

olive ember
#

U are setting up a pc to be able to powercycle some "servers" that could be crashing, and this pc is already running in emergency/recovery mode ?
Intersting. πŸ™‚

ripe saffron
#

question, i'm transferring my HA to a VM. I still have my old HA open to move one by one to make sure its stable. The issue is I keep getting IP ban on my new HA. I cant figure out why?

humble mirage
#

And ... ? What IP is banned?

gloomy orbit
#

I'm wiping my proxmox host and installing a new major version. Are there any differences in restoring from HA backup than rebuilding a VM from a backup? I.e. do you risk database corruption/mismatches etc if you just pull a restore of the whole VM?

hasty osprey
#

Faster than a GUI restore too

raven forge
dapper sedge
#

you could setup the HA container with a Macvlan network so that HA is a device with a physical IP address on your network

gloomy orbit
#

Is it possible to just restore a backup if you're moving from HAOS -> Docker install? Obviously without the addons

raven forge
#

Sure, but you have to extract the files yourself.

sudden junco
#

Trying out HASS OS on a device not at my location - fresh install. I'm getting hung-up on onboarding page everytime, is there a way to RDP into the CLI out of the box or gotta wait until I can get to the box again?

raven forge
#

No. SSH is not enabled by default.

sudden junco
#

Ah nah seems I want the addon, but how do you install the addon if you can't get to admin panel? Just drop it in SD on another device?

sudden junco
#

I went to get the device - I re-installed it and on onboarding screen it can't create the first user. Get a generic 500. I checked the logs on the device but it doesn't really mention anything about what is going on.

I looked at the file structure and see no way to manually drop an SSH file like mentioned in a couple links, nor a way to install SSH without logging in. How do I troubleshoot the issue?

raven forge
#

How did you check the logs? Did you create a user with uppercase characters?

sudden junco
#

I can't create a user on the site - And if I reboot it no longer triggers onboarding and shows an empty array for users.

I used HA CLI to check the cli, and then used one of the other log commands - I forget which. It was more device logs if I remember.

raven forge
#

The common ones are ha host logs, ha supervisor logs and ha core logs.
I'd try ha os datadisk wipe to reset and see if it works afterwards.

sudden junco
#

Thanks I'll try ha os datadisk wipe tomorrow - I was looking for something like that before I finally just brought the card back with me

sudden junco
#

So I finally got a fresh install, but getting the exact same error when I try to create an initial user. 500 on api/onboarding/users
HASS OS on a Pi, local network connect.

sudden junco
#

I decided to add a last initial to 'Name' and suddenly that let it pass. Seems there is just no error validation or UI response if the user fails to create.

karmic hemlock
#

hello everyone, i am new to home assistant and am trying to install it on my pi 400. I followed all the steps from the page but the home assistant cli is not starting. I have tried 'login', ctrl+alt+f2 with root as login helped me get to the os but i cannot start the home assistant cli with 'ha', it says i do not have hassio_cli so i tried 'docker pull' it said it cannot verify the certificates, i tried changing dns but that did not work. i dont know what to do

viral scroll
#

Install Generic x86-64

gloomy orbit
#

roses are red
violets are blue
my backup restoration is on minute 52

#

For a larger installation, this process can take about 45 minutes. We're entering uncharted territory

raven forge
#

How big is your backup?

gloomy orbit
#

1.8 gb

raven forge
#

I don't see why it would take this long. Even on the slowest pi.

gloomy orbit
#

I suppose the ip of the initial machine doesn't need to be the same as set in the backed up version?

#

I restored from the new login screen. Maybe I should create a user and restore from settings, the old fashioned way

raven forge
#

Depending on how it was configured the backup would probably restore that. Not sure if a HAOS backup includes OS settings.
It's recommend to restore through samba if your backup is larger than 1G. At least it was, I can't find it right now. Maybe it was removed or I was hallucinating.
I'd try to open HA in a new tab and see if it's accessible. Otherwise you can run ha supervisor logs -vf from the CLI to follow the progress.

gloomy orbit
#

Thanks, not really sure what I am looking for in those logs though

#

Latest log message was 5 mins ago, Updated Home Assistant API Token

raven forge
#

Just if something happens mostly.

gloomy orbit
#

Seems to be running some checks, which I suppose is a good sign...?

raven forge
#

Hmm. Apparently a static ip setting will not be restored. At least I can't find my static ip referenced in the full backup's files.
As long as something is happening that's useful.

gloomy orbit
#

Right

#

An idea: I have also mapped the ip in my router. But is it the ha "container" (it's HAOS but in lack of a better word) or the proxmox VM that is mapped - i.e. does the mac adress belong to the proxmox VM I'm running HA in, or HA itself?

I suppose what I mean is that I suspect that this could cause an ip conflict

raven forge
#

The MAC belongs to the virtual NIC of the VM so that won't change even if you install an entirely different OS.

gloomy orbit
#

Right. And as this is a new VM, the ip that is set to static in HA is already assigned to my old VM in my router. So I suspect this is causing a conflict then

#

I'll just release it and start over I think.

#

I don't know if it's relevant, but the dev tools console at the restore page is giving me a 400 to :8123/api/hassio/backups/52ff9823/restore/full:1

raven forge
#

I'm not sure a static ip is restored. I think it will just get a new ip via DHCP by default.
Btw if you use proxmox VE I'd recommend you make use its backups/snapshots as well. Both have their pros and cons.

gloomy orbit
#

I have a proxmox backup as well, trying that next.

gloomy orbit
# raven forge I'm not sure a static ip is restored. I think it will just get a new ip via DHCP...

😬

vma: restore failed - short vma extent (3706624 < 3801600)
/bin/bash: line 1: 69518 Exit 1                  zstd -q -d -c /var/lib/vz/dump/vzdump-qemu-200-2024_07_27-11_05_55.vma.zst
     69519 Trace/breakpoint trap   | vma extract -v -r /var/tmp/vzdumptmp69509.fifo - /var/tmp/vzdumptmp69509
  Logical volume "vm-200-disk-0" successfully removed.
temporary volume 'local-lvm:vm-200-disk-0' sucessfuly removed
  Logical volume "vm-200-disk-1" successfully removed.
temporary volume 'local-lvm:vm-200-disk-1' sucessfuly removed
no lock found trying to remove 'create'  lock
error before or during data restore, some or all disks were not completely restored. VM 200 state is NOT cleaned up.
TASK ERROR: command 'set -o pipefail && zstd -q -d -c /var/lib/vz/dump/vzdump-qemu-200-2024_07_27-11_05_55.vma.zst | vma extract -v -r /var/tmp/vzdumptmp69509.fifo - /var/tmp/vzdumptmp69509' failed: exit code 133```
raven forge
#

Oof.

gloomy orbit
#

Well I have older HA backups, guess I'll be trying them.

raven forge
#

You can try zstd -t /path/to/backups/dump/*.vma.zst to see which ones aren't broken.

-t, --test
Test the integrity of compressed files. This option is equivalent to --decompress --stdout > /dev/null, decompressed data is discarded and checksummed for errors. No files are created or removed.

gloomy orbit
#

Thanks, but I only have one proxmox backup sadly, multiple ones from "inside" HA though.

#

Intresting - Backup was made on supervisor version 2024.07.0, can't restore on 2024.06.2. Must update supervisor first.

#

No error in the GUI thoguh

raven forge
#

Just out of curiosity, was the qemu agent enabled during the backup? I don't expect this to cause this specific issue but still.

gloomy orbit
#

Tbh not sure. It was on whatever is default

raven forge
#

Ah. Because if it isn't your database and certain files might be saved in a corrupted or half-written state. The vma and zst should not be affected though.
You can see if it's enabled in the VM's Summary/Options page.

gloomy orbit
#

Right, I did the upgrade from proxmox 7 -> 8 with a fresh install, so the vm is not existing anymore (as the backup is corrupt lol)

#

But the backup was run with the full stop VM option, so it sounds strange that that would be the case.

raven forge
gloomy orbit
#

Yeah I have a computer laying around for testing. Guess it's time to make it a PBS box.

raven forge
#

I use it in a VM like a mad man πŸ˜„

gloomy orbit
#

...inside your host? πŸ˜„

copper steppe
#

Have fun!

raven forge
#

I use one of my nodes as a storage node and that's where I also run PBS on.
The PBS VM is backed up via PVE backups btw.

gloomy orbit
#

Interesting, maybe I'd try the same

#

However right now I just want to be able to turn off my lights and go to sleep...

raven forge
#

Understandable.

gloomy orbit
#

Well, so far so good. I had to create a temporary account in the new install, enable the beta updates (isn't supervisor 2024.7 stable now? https://github.com/home-assistant/supervisor/releases/tag/2024.07.0) to see the supervisor update, run the update, then upload and restore the backup. Not done yet but the logs seems to be a lot more happy now.

raven forge
gloomy orbit
#

Ahhh it worked!

#

Thanks for you support mate.

raven forge
#

You did most of the work but glad it worked.

gloomy orbit
#

Still interesting why the proxmox backup corrupted though. I'll give you forum link a read tomorrow

karmic hemlock
fluid yoke
#

I have installed the latest home assistant in an old i3 pc with an SSD. Everything seemed to go well but on booting I get "Failed to start containerd container routine" and then it will just sit so at "waiting for cli to start". Does anyone have a solution or know what the failure is?

wintry breach
#

Hey, i want to setup remote access ha. Tried the portforwarding method, didnt work because of some unknown issue with my router, so i was wondering if someone could suggest other free to implement methods...

raven forge
#

Tailscale.

daring sun
# karmic hemlock anyone please?

@karmic hemlock Have you any failures on boot? I get a failure to start containerd contained routine. I also just get waiting for cli to start. I don’t know if it has anything to do with the routine not starting.

karmic hemlock
daring sun
#

@karmic hemlock I’ll take another look at mine but it’s the container routine that’s the only failure I can see on my HA boot.

karmic hemlock
raven forge
raven forge
karmic hemlock
warm hound
#

Anyone have a tip how I can fix this error?

Supervisor:    Connected
Supported:    Processing...
Healthy:    Processing...```

This ad has been there for about 10 hours, and I've also had the problem over the last few days that homeassistant is unavailable from time to time for a few minutes. Runs on dedicated x86 hardware with SSD
wintry breach
viral scroll
wintry breach
steep thistle
#

Hi, I installed Home Assistant on Raspbery pi 3 B. I use 8 gb micro SD card I already have. Everything seem work fine. But after some hour of "inactivity" it stop working. Led of the RJ45 port turn off and there is no more signal on HDMI. Home assistant won't respond. The only life sign of the raspbery is solid red led, no more.
Also I don't have the genuine power supply, so I've "Undervoltage detected!" messages. I don't want to order PSU without be sure what will fix that.

Is anybody know how can I solve the problem ?

Thank you.

proven grove
steep thistle
#

So you're almost sure that my issue could be fixed with geguine PSU ?

open fulcrum
#

especially with the undervoltage detected messages

steep thistle
#

95% is enought to take a chance buying one to get stable home assistant.

proven grove
#

I mean, you should have a proper power supply anyway. Not sure why that's even a question. lol

open fulcrum
#

you might already have one lying around

#

especially from any old phones bought around 10 years ago

#

afaik

steep thistle
#

I've only 5v 2a PSU currently. The person that give me the pi3 lost the PSU.

#

is it could be better choice to just bought Home Assistant Green than try to get stable experience with old pi3 ? If I've to invest it's for a stable solution.

#

On Home assistant website they say that pi3 is just ok no more to run it.

proven grove
celest quartz
#

Hello, Restoring from a 11 gb backup on a new NUC, been 9 hrs now and still going. This normal?

tender wren
celest quartz
#

It gives the onboarding screen 😬

#

I have it connected to hdmi and have cli access.

#

Anything i can do to see anything has progressed?

tender wren
#

I would not recommend to upload such a big backup file via the onboarding screen.
Better way would be to do the onboarding and uploading the backup either via the backup page or with a samba connection - and triggering the actual restore process from the backup page after a successful upload.

tender wren
celest quartz
#

Ahh this would be good.

#

I can try that.

#

Not sure what to do with the current one

#

Just stop it ?

#

Or just reboot the system via cli

tender wren
#

A reboot should stop the process

humble scaffoldBOT
warm iron
#

ope lol

#

I'm honestly just baffled, because this seems to've cropped up out of nowhere β€” I've even gone back to older stable builds of the image (latest, stable, 2024.4.* and 2024.6.2) and the same error crops up on a totally fresh HASS setup. It felt initially like the container was (somehow) running in x32, but running uname within the container confirmed it was running on the same x86_64 arch as the OS. Even running the container without any volumes mounted to the OS yields the same error.

#

This is more of a sanity check for myself to definitively make sure it's something on my end and not, say, an issue with the python dependency being pulled into the container processes

tender wren
#

What installation type are you using?

warm iron
tender wren
#

Please share the compose file

warm iron
#

for sure! Here's the compose:

services:
    home-assistant:
        container_name: hass
        privileged: true
        # restart: unless-stopped
        volumes:
            - /var/run/docker.sock:/var/run/docker.sock
            - /data/home-assistant:/config
            - /etc/localtime:/etc/localtime:ro
        network_mode: host
        image: ghcr.io/home-assistant/home-assistant:stable

I've been trying to pare it down to the bare minimum just to get something running successfully

#

Docker itself is being run as root, and the /data/home-assistant directory is getting populated successfully with the correct chown.

#

I'd really just be happy with a quick sanity check of "yes, on another machine, stable is setting up correctly and this is something on your end" just to make sure i'm not going crazy lol

tender wren
#

I'm not sure if the additional spaces in your compose file cause an issue.

# 
services:
  homeassistant:
    container_name: homeassistant
    image: "ghcr.io/home-assistant/home-assistant:stable"
    volumes:
      - /data/home-assistant:/config
      - /etc/localtime:/etc/localtime:ro
      - /run/dbus:/run/dbus:ro
    restart: unless-stopped
    privileged: true
    network_mode: host
warm iron
#

The compose is processing correctly and the HASS container is starting so I don't think that'd be it, but I can give it a go just to weed out the possibility

humble scaffoldBOT
warm iron
#

(this message-to-file dealio is actually pretty nice for condensing long messages β€” props for that)

#

The reason I bring this up is because the error seems to be in the Python running within the container and not anything directly on the OS, which feels symptomatic of the container build.

#

Even running the process as the absolute bare-minimum yields the same error, with the compose file:

services:
  home-assistant:
    container_name: hass
    image: ghcr.io/home-assistant/home-assistant:stable
tender wren
#

Try the compose file I posted without any changes

#

No renaming, no cutting entries

humble scaffoldBOT
warm iron
#

Although! In a zesty development, the process does work temporarily after the restart policy kicks it off again β€” however, it crashes with the same error once the UI is fetched from the browser.

humble scaffoldBOT
#

@wet remnant I converted your message into a file since it's above 15 lines :+1:

πŸ“Ž message.txt
wet remnant
#

Google/Github can be accessed normally on Windows, no linking issues.

celest quartz
proven grove
#

Also, what’s the docker command you’re using?

celest quartz
#

any idea what's going on?

#

Also not sure that this is related to the problem but it's asking for a password which is weird because the backup was not made with the password, so I have to put this as a password "\"

tender wren
#

Might be a corrupted backup file

celest quartz
#

it talks about securetar.SecureTarFile maybe I should make a password for it?

tender wren
#

A password is not necessary for a backup

#

It is an option

#

Does the restore process start without entering a password?

celest quartz
#

no it does not start without that password

tender wren
#

In that case it either has been creates with a password (maybe by mistake) or the backup file is corrupted.
I would try to re-upload the file. Maybe something went wrong during the upload.

urban tusk
#

Hey guys, I'm getting all kinds of weird effects in my HAOS. It's not exactly crashes, more like unresponsive, half-rendered or not rendering UI. Can somebody please help me debug that?

#

Sometimes it recovers, other times it gets so bad I have to power-cycle the entire machine.

#

It's not really consistent either, sometimes it doesn't happen for days, then it happens twice in an hour.

#

It feels like it's getting more frequent over time, but I have no consistent proof of that either.

tender wren
#

Restart HAOS in safe mode and observe, if those issues still occur. If not, a custom integration or card might be the issue.

urban tusk
#

I'm looking at the logs right now, and I see some weird messages about HACS blocking the event loop. Does that sound like a lead?

humble scaffoldBOT
#

If you're having problems with your updates to your configuration:

urban tusk
#

What was that trick that gave you root SSH access to the host OS? I'd like to run some top and see if anything goes haywire in the background?

#

btw, I reloaded the page and it started working again. Looks to me like some api the UI uses is sporadicly unresponsive.

tender wren
#

I'd recommend to go the safe mode route first to rule out custom integrations or cards.

urban tusk
#

I've got it running in safe mode now.

#

Sorry if I'm a little frustrated. I'm a 25 year veteran of Linux and sysadmining and all I get to do is to essentially turn it off and on again. There's GOT to be something more

#

useful I can do

#

Ok, it's not the custom integrations. It's running in safe mode and shows the exact same behavior.

tender wren
urban tusk
#

preparing the USB drive as we speak.

tender wren
#

You can also use the Glances add-on to get some insights

urban tusk
#

Are there any known issues with nvme ssds on rpi5?

#

It looks to me a lot like a bottleneck issue.

#

If for some reason some read calls were blocking, that would explain why it gets into trouble.

urban tusk
urban tusk
#

I'm beginning to suspect strongly that this is a web server or web api issue, because even though the UI is entirely locked up atm, the automations are triggering normally.

#

That is, at least some of them are.

civic mango
#

SkyConnect: unsupported firmware | ssh to core system

jagged sapphire
#

I'm trying to convert from HA core to HAOS. I took a backup of core, spun up HAOS In a vm and tried to restore from backup and it seems to never complete even after 4 hours. If i reboot, i can no longer connect to the website for HA. Any ideas?

dapper sedge
jagged sapphire
#

yes and no. i want a fully reproducible system in case of an issue. so i can just spin up a new vm, install, restore from an older backup and be up and running. with ha core im maintaining node red, mqtt, etc all in separate dockers and have to remember to update them, back them up, and then if i need to restore i have to repeat it all

dapper sedge
#

gotcha. I also run in docker and have an MQTT and Frigate container running, no node red. I personally keep volume mappings from my host into the containers and have rsync cron jobs to back up all the configs and important files when needed. For me, that solution is fully reproducible and quick to spin up when needed

#

but I do manually update everything

dapper sedge
jagged sapphire
#

why yes i do, i didnt know it would cause issues with the backup?

#

im certain ive done backups/restores with HACS in the past

dapper sedge
#

well I don't know if it would, I just suspect custom software installed on HA could cause issues with converting to a clean setup

dapper sedge
humble mirage
wind osprey
#

Hey! My ha inside docker is not starting anymore (the docker is up but a sensor is stopping homeassistant inside)

2024-07-29 09:03:14.157 ERROR (MainThread) [homeassistant.components.template.template_entity] TemplateError('ValueError: Template error: float got invalid input 'unknown' when rendering template '{{ states('sensor.sensor_zelt_2') | float - 25 }}' but no default was specified') while processing template 'Template<template=({{ states('sensor.sensor_zelt_2') | float - 25 }}) renders=4>' for attribute '_attr_native_value' in entity 'sensor.rel_luftfeuchte_mitte_25'
2024-07-29 09:03:14.164 WARNING (MainThread) [homeassistant.components.ssdp] Could not set up UPnP/SSDP server, as a presentation URL could not be determined; Please configure your internal URL in the Home Assistant general configuration

any idea how to fix / find the files to correct? I don't even understand why it's crashing because of some values β€” or is the upnp thing mentioned as severe?

raven forge
#

Try grep -sR "sensor_zelt_2" /pathtohaconfigdirectory to find where the template is referenced.
This should not prevent HA from starting though.

humble mirage
#

What does the command for starting the container look like/the compose file entry?

wind osprey
#

tbh I don't have a compose... but I removed the old container and started it again with this command:

run -d -v /etc/localtime:/etc/localtime:ro \
  --name homeassistant \
  --privileged \
  --restart=unless-stopped \
  -v /home/jan/.homeassistant:/config \
  -v /run/dbus:/run/dbus:ro \
  --network=host \
  ghcr.io/home-assistant/home-assistant:stable

I just added the -v for the time (because it was way off... sorry it made the cmd very chaotic πŸ˜€)

humble mirage
#

You should use compose, but that at least looks correct

#

What hardware are you running on?

wind osprey
#

Raspberry Pi 5

humble mirage
#

Also what Linux distro?

wind osprey
#

Linux MICROLABS-P5 6.1.0-rpi8-rpi-2712 #1 SMP PREEMPT Debian 1:6.1.73-1+rpt1 (2024-01-25) aarch64 GNU/Linux

humble mirage
#

Actual Debian? Pi OS? Something else?

wind osprey
#

Yeah Raspbian I guess?

#

PatchBox OS it's called iirc based on raspbian?

wind osprey
# raven forge Try `grep -sR "sensor_zelt_2" /pathtohaconfigdirectory` to find where the templa... 
grep -sR "sensor_zelt_2" .οΏ½οΏ½οΏ½οΏ½οΏ½οΏ½οΏ½οΏ½οΏ½οΏ½οΏ½οΏ½οΏ½οΏ½οΏ½οΏ½οΏ½οΏ½οΏ½ 09:03:14.152 ERROR (MainThread) [homeassistant.helpers.event] Error while processing template: Template<template=({{ states('sensor.sensor_zelt_2') | float - 25 }}) renders=2>
./home-assistant.log:ValueError: Template error: float got invalid input 'unknown' when rendering template '{{ states('sensor.sensor_zelt_2') | float - 25 }}' but no default was specified
./home-assistant.log:homeassistant.exceptions.TemplateError: ValueError: Template error: float got invalid input 'unknown' when rendering template '{{ states('sensor.sensor_zelt_2') | float - 25 }}' but no default was specified
./home-assistant.log:2024-07-29 09:03:14.157 ERROR (MainThread) [homeassistant.components.template.template_entity] TemplateError('ValueError: Template error: float got invalid input 'unknown' when rendering template '{{ states('sensor.sensor_zelt_2') | float - 25 }}' but no default was specified') while processing template 'Template<template=({{ states('sensor.sensor_zelt_2') | float - 25 }}) renders=4>' for attribute '_attr_native_value' in entity 'sensor.rel_luftfeuchte_mitte_25'
#

How can I remove this script entity? I don't need it anymore anyways since this was just a hack for a humidity sensor that was a bit off

raven forge
#

I can't really see where it should be hiding. grep -slR /home/jan/.homeassistant should show all files where it's referenced.
Might be coming from somewhere else?

wind osprey
#

is it safe to edit in there?

raven forge
#

You generally don't want to fiddle with those files. I assume you have a backup so have at it. I'm not convinced this will fix the not starting part though.

wind osprey
#

and the db

humble mirage
#

Then you set up the template sensor in the UI

#

And so ... ||that's where you remove it||

wind osprey
#

but I can't connect to the UI. Have I mentioned that? πŸ˜€

humble mirage
#

Sure, but that template isn't the cause of your problem

wind osprey
#

okay... hmh

dc26f358f463   ghcr.io/home-assistant/home-assistant:stable   "/init"                  59 minutes ago   Up 57 minutes                                                           homeassistant

so container is running... but UI is not reachable. Hmm.

humble mirage
#

You didn't change the port in configuration.yaml?

wind osprey
#

not ever

humble scaffoldBOT
#

@wind osprey I converted your message into a file since it's above 15 lines :+1:

πŸ“Ž message.txt
humble mirage
#

You're remembering to use https://?

wind osprey
humble mirage
#

Well, you've added the ssl_ lines, so you have to use https:// even locally

#

http://192.168.0.42:8123 won't work any more

wind osprey
#

oh hm

#

man... thank you

#

removed the http and cert parts and it came back up immediately

humble mirage
#

It always was up

#

You just weren't using https://

wind osprey
#

yeah in the app I meant πŸ™‚

#

I want to tho πŸ₯²

#

so difficult to expose ha to the internet / make it connectable... it never answers my calls

humble mirage
#
  1. Pay for Nabu Casa
  2. Set it up manually with a reverse proxy
  3. Set it up manually directly, as you were
wind osprey
#

anyhow my learnings: MainThead Errors don't mean ha is in a crashed state

wind osprey
humble mirage
#

"hidden"

wind osprey
humble mirage
#

Somebody needs a better browser/their eyes tested

wind osprey
#

oh yeah... last time I looked I couldn't find any numbers while scanning the document and my brain saved that information - that's my excuse πŸ˜‚ maybe make them bold

#

I'm a super average human so if it happens with me it happens for a few thousand others as well πŸ˜…

humble mirage
#

That's not gonna help you much πŸ˜›

#

Everybody else can spot it

wind osprey
#

as I said, I'm the most average dau there is - you're in the 1% of the smartest people alive πŸ˜… probably, take that as a compliment

oak wyvern
#

greetings fellas, currently have a beelink s12 pro mini pc, i need a conveter adapter, from m.2 sata3 2280 ssd to M.2-2230 (E-key)

#

i would need an adapter that is able to install it horizontally not vertically

#

feel free to @ me

umbral grail
#

Hallo to all, I have installed the qcow2-Image on a proxmox server, the machine is booting well. I can ping the virtual machine and got a result in the browser with to observer url (port 4357). The problem is that there is no response on the HomeAssistant-URL (port 8123). Does anybody have some hints to solve the problem?
Thanks, rmb

raven forge
#

Try ha supervisor logs and ha core logs.

umbral grail
oak wyvern
#

greetings fellas, anyone know of a m.2 2280 -> m.2 2230 e-key converter

#

considering installing a google coral tpu onto a mini pc

celest quartz
#

thanks for the help πŸ™‚

oak wyvern
#

darn, there does not appear to be some sorta converter from m.2 2280 to m.2 2230 e-key

dapper sedge
tender wren
#

Do you have the 2230 coral or a 2230 slot?

oak wyvern
oak wyvern
#

i would need to go from m.2 2280 to m.2 2230 e-key

tender wren
#

Otherwise there would be a 2280 Coral module with b and m key

oak wyvern
#

hmm i think the issue with that is that the dual is only m.2 2230? correct

dapper sedge
oak wyvern
#

and the m.2 2280 is only 2 TOPS

#

i could be mistaken

oak wyvern
dapper sedge
#

I haven't tried it but I'd assume it would yield similar results. What CPU do you have in your mini PC?

oak wyvern
#

its a beelink s12 mini pro

#

Quad-core 3.40 Ghz (max 3.40 Ghz) Intel Processor N100, 16gb ram, 500 ssd

#

its a puppy

#

would it be beefy enough to use this openvino?

dapper sedge
dapper sedge
tender wren
#

This is a bit too off-topic for this channel. Please start a topic for that or head over to #cameras-archived

oak wyvern
little atlas
#

My Pi4 HA slowed down a lot, was thinking it was a bad SD card, so I plugged it in my laptop, and saw it were using all its space. But the partitions were leaving 58,9 GB un-partitioned.
From what I read HA changes the data partition when it boots, right? So it's not what's causing the slowdown? Slowdown happened when I set up HA Cloud (And feels just as slow via local IP) πŸ€”

urban tusk
#

fI'm beginning to suspect my troubles with the HA UI have more to do with the browser or its interaction with the server than the server itself.

little atlas
#

Oh right, forgot to say that I already assured it wasn't an browser/app issue. When I say slow I also mean switches being slow to talk with lights via Zigbee, like, in the seconds responds time, just like frontend ui loading

north radish
#

I'm trying to get HA running in a new VM using Virtualbox (VB 7.0.2) on my Intel Macbook Pro (Sonoma 14.2.1) using haos_ova_12.4.vdi. I setup the VM according to the instructions on the Install page and have EFI turned on.
When starting the machine, I'm left with a blank screen, no disk activity, and I'm seeing messages like the following in VBoxSVC.log: https://pastebin.com/3T9m4Hwz I.e. the machine is not mutable. I haven't had this issue before.
Would anyone know what's wrong?

north radish
# north radish I'm trying to get HA running in a new VM using Virtualbox (VB 7.0.2) on my Inte...

Ok, turns out that I wasn't looking at the right log file. In the machine's VBox.log file (after deleting existing log files) I'm getting repeated messages:
00:00:53.133637 ERROR [COM]: aRC=VBOX_E_VM_ERROR (0x80bb0003) aIID={GUID} aComponent={DisplayWrap} aText={Could not take a screenshot (VERR_NOT_SUPPORTED)}, preserve=false aResultDetail=-37, and nothing else. The VM seems to be stuck but I'm not sure what's happening. (I'm trying haos_ova_12.3.vdi to see if it has anything to do with the new version.)

north radish
north radish
winged bramble
#

Hi,
How can we add a route on HAOS that uses a specific interface of HAOS to target a specific subnet ?
Using the nmcli commands it adds routes, but always on the default interface...

I'm looking at how I can use nmcli and persist the route to have a route like:
192.168.5.0/24 via 192.168.10.1 dev end0.10 (manually added using ip route add)
Using nmcli I only get such route:192.168.5.0/24 via 192.168.10.252 dev end0 proto static metric 100 (this ws created through nmcli, but it is assiged to incorrect interface)

humble mirage
#

Why not have your router handle that?

#

Or do you have multiple interfaces set up on HAOS?

winged bramble
#

so I think that when I added a route using the connection name, it was actully selecting a wrong connection and therefore not routing over proper interface.
I did a cleanup of those connections and adding a route now does indicate the right interface to be used.

lost axle
#

How do i make HAOS realize it's platform has changed?
It does see the new RAM but it still thinks it's running on a Pi4 https://i.vgy.me/gyS3Kc.png

raven forge
#

You should re-flash and restore a backup. There's a reason there's so many different builds. What did you change to?

lost axle
#

just from pi4 to 5

#

honestly if theres no other way than reinstall i'll probably just let it continue thinking it's on a pi4

#

so far everything seems to be working as before anyway

tender wren
#

The RPi5 has its own image for a reason

crude inlet
#

Pi5 image for the Pi5's hardware

lost axle
#

okay all the guys that say they agree, add me, lets voice call and you do it for me via vnc πŸ™ƒ

crude inlet
#

Well, you create and download a full backup first.

#

And then you flash an SD with the image for Pi5

lost axle
#

nah, dont walk me through it, thats no better than reading the docs :D

tender wren
lost axle
#

you can't be bothered as much as me, so we're even again

tender wren
#

You heard our recommendation. Now it is up to you to choose your poison

lost axle
#

i have already chosen like 15 minutes ago

#

i moved on to other things in the migration chain of Pis

#

now onto migrating motioneyeOS from a Pi3 to 4 that magically is free now :)

grand pivot
urban tusk
#

Can anyone give me a reason why nslookup on two different machines would return a IP address for a certain hostname, but chrome would claim the domain is unreachable and nslookup would claim it can't resolve the hostname?

#

Does that make a nickel's worth of sense?

#

Oh btw, if I go via the IP address, the host is up and responding

raven forge
#

Please share what you saw.

urban tusk
#

Never mind, clearing the OS dns cache seems to have fixed it for now.

#

I think I need to set up a proper DNS in my network. Those rudimentary router-supplied DNS are just crap.

raven forge
#

Unless you need specific features and have a okay router it should be okay.

urban tusk
#

Depends on whether or not you call having no zombie DNS entries from long-daed DHCP leases decent support for static IP a "special features" which I do not

urban tusk
#

I think I've finally got everything working again. I can tell you I was just about ready to chuck the whole thing kit and kaboodle.

dull tangle
#

what's the latest OS that works on a fresh install with the rpi4? 12.4 doesn't load the preparing page at all, 12.0 loads but has errors.

proven grove
gray bobcat
#

Hi can someone help me out? 2024.7.4 isnt able to update. When Ipress the update button its tying to update an then zb grafana and lots of other integrations do not work

#

I have to power on my intel nuc and the 2024.7.3 is back someone any idea

raven forge
#

Tail the supervisor log via SSH while you update.

ha supervisor logs -vf
gray bobcat
#

ok no idea what you are talking about

raven forge
#

Run this command in the SSH addon, then trigger the update and see what is logged.

gray bobcat
#

and the ?

#

n

raven forge
#

If you can't make sense of what is logged share it.

humble scaffoldBOT
#

Please use a code share site to share code or logs, for example:

Please don't use Pastebin, since it can randomly add spaces to the main view. Please also don't share text as images since it makes it harder for people to help you. Remember that others may have colour blindness, impaired vision, etc.

gray bobcat
#

NO CLUE HOW TO GET THE LOG OUT OFF THE TERMINAL'

dull tangle
raven forge
gray bobcat
#

does not work

raven forge
#

Very helpful.

gray bobcat
#

hey i am trying my best but cant get the info out of the terminal

raven forge
#

I gave you three options. There's nothing else I can tell you.

dull tangle
#

12.2 also has a Can't find Supervisor network error like 12.0 did. ugh

gray bobcat
#

I dont blame you

#

I ll try tomorrow thx for your help I have to get more deeper inside ha. Otherwise its hard to follow

dull tangle
#

Resolved, turned out to be a NTP server error the logs weren't telling me about

autumn niche
#

Hey all, I’ve been digging around for a few hours now and can’t seem to figure this out. Does HA green work with thread/matter? I have the ZBT-1 plugged in, went through the installation I thought was needed, have the matter, thread, and thread border integrations installed on the device but adding a new matter device (Eve motion sensor) won’t work. I’ve tried the door sensor, motion sensor and a nanoleaf light. I’ve scanned the QR code maybe 30 times, won’t connect. I’m pretty new to the device but it all seemed pretty user friendly to set up. Anyone know what I might be missing? Wondering if I didn’t install something or enable something. I did have to turn on matter, which was not on by default.

placid sage
#

Hi. I'd like to install HAOS on an old laptop with 8GB RAM. Would it be better in the long term to install it bare metal, or in a VM Proxmox guest?

crude inlet
crude inlet
placid sage
#

Hm, actually, I'll wanna replace that laptop with a steamdeck once steamdeck2 comes out. First gen (and oled) have 16G RAM and great power efficiency.

raven forge
#

It changes whether HAOS can be installed bare-metal or not as it requires EFI boot. A VM can use EFI firmware even if the host's hardware doesn't support it.

placid sage
#

Alright, thanks

quartz plover
#

I have just performed upgrade of ha from .2 to .4 I'm just stuck in startup with "RecursionError: maximum recursion depth exceeded", kind of really bad luck. But since its stuck in startup sequence, I cannot even do restore or install older version using ha cli. Any idea how to proceed here?

humble mirage
#

What are the actual version numbers you're hinting at there?

quartz plover
#

just a minor version update 2024.7.2 to 2024.7.4

rare yew
#

having some trouble with frequent restarts on fresh install. Also having a hard time installing certain add-ons. Mostly vanilla.
zwave js ui, matter, vesync, music server (beta), reolink, Which dropdown value in the logs is the best to try spot the issue?
Its hard because the pages dont load, or it restarts on me.

humble mirage
#

I'd start with host and supervisor

rare yew
# humble mirage I'd start with host and supervisor

supervisor: only has the startup infos since the last crash.
home-assistant: [homeassistant.components.recorder.util] The system could not validate that the sqlite3 database at //config/home-assistant_v2.db was shutdown cleanly
host: time="2024-07-31T11:02:23.427236640Z" level=error msg="stream copy error: reading from a closed fifo" ( a few like this )

nothing really standing out.
corrupt memory on SD Card?

raven forge
#

Too little information to tell.

south dirge
#

Hi Everyone, new here! I am installing the home assistant boot loader to my SD but when it is complete it creates a bunch of partitions and gives me quite a few errors about formatting them all. I also cannot boot into HA after plugging it into my Pi5. Please help

crude inlet
#

Yes many partitions are created, and you should ignore all offers to format them

south dirge
#

I do, but it still doesn’t boot correctly on the pi 😦

crude inlet
#

Are you watching it fail to boot (keyboard and monitor connected to the pi?)

south dirge
#

No, it is solid green light for a little, blinks briefly, then fully red and never changes.

#

And I can’t access it through the web browser

crude inlet
south dirge
crude inlet
#

Yes, I’ve reflashed it probably 5 times

finite kite
#

Hi there, I'm trying to install Home Assistant OS on my Truenas Scale in a virtual machine, to do this I followed this tutorial: https://community.home-assistant.io/t/installing-haos-in-a-vm-on-truenas-scale/511393

Everything went well until I tried to access the web interface, I received the following message:

This website cannot be accessed. Page 192.168.1.137 refused the connection.

I have tried to access with said IP and with http://homeassistant.local:8123/, with this second method I get the message:

This website cannot be accessed Check for a write error in homeassistant.local.

If you could help me with this problem I would appreciate it

olive ember
#

Looks lke your VM cannot be reached. Check your network settings.

finite kite
# olive ember Looks lke your VM cannot be reached. Check your network settings.

Thank you very much for answering, I have been thinking about it and I think maybe it is due to this:
My Truenas has an IP with a /24 subnet mask and the Home Assistant IP has the same subnet mask ,this gave me problems in the past with other applications that I have installed on Truenas, changing its subnet mask to /32 solved it
Do you think this could be the problem? And if so, do you know how to change it in Home Assistant?

humble scaffoldBOT
#

@alpine narwhal I converted your message into a file since it's above 15 lines :+1:

πŸ“Ž message.txt
alpine narwhal
#

Wish I knew that, I would have tried to make it shorter instead πŸ™‚

brazen herald
#

Try removing the internal and external URLs from the YAML and configuring it in the UI.

alpine narwhal
#

thanks, just a sec πŸ™‚

#

should I leave that http: section I added? or remove that too?

brazen herald
#

The http section is necessary. As you're doing a reverse proxy you'll need those two config parts.

alpine narwhal
#

ok... so in the UI, I entered the addresses correctly but when I hit Save the external (labelled Internet) field deleted itself (went empty)

brazen herald
#

For me (granted I'm using Caddy), all I've specified in the proxy is ip.add.re.ss:8123, that's without the protocol and without the trailing slash.

brazen herald
alpine narwhal
#

yes

brazen herald
#

Do you see anything in the logs?

alpine narwhal
#

it's giving an error for a Docker IP... so I'll add that to the proxy lists...

brazen herald
#

Ahh yes. You might want to copy the docs there and add the whole subnet of docker IPs

alpine narwhal
#

restarting again now

#

it loaded

#

Thank you so much

brazen herald
#

No problem, glad to have helped!!

alpine narwhal
#

spoke too soon

#

when I entered my username/password I got Unable to connect to Home Assistant and a countdown timer

#

now I can't even go back the url to even check anything, it autoredirects to that countdown

#

the logs make it look like I had invalid user/pass but I stored them in my password manager so I know they are correct... could it be because I'm still logged in via IP:Port?

#

nevermind... sorry to bother you... I needed to turn on Websockets in NPM

olive ember
latent thistle
# 
ha core update --version 2024.7.3
Processing... Done.

Post "http://supervisor/core/update": context deadline exceeded (Client.Timeout exceeded while awaiting headers)
#

how to fix this?

wintry yoke
#

Hello!
How can I access the Home Assistant frontend code?
I tried to follow the tutorial with Docker and got a JS heapout of memory error... I also tried the command line and it gave the same error. I did the set node_options command but it didn't change anything.
I'd like to use the code and change it according to my needs.

I hope someone can help me!
Thank you.

humble mirage
#

Mostly you don't

#

What are you trying to do?

sly sluice
#

Hello, I reinstalled Home Assistant on an old PC but everytime the PC boots it says no boot device. Can somebody help me?

raven forge
#

How did you install it? Does that PC support EFI booting?

humble mirage
#

Hammer?

sly sluice
raven forge
#

What steps did you follow to put HAOS on the boot device? What boot device is it? Does it show in the BIOS?

sly sluice
#

I installed it over Ubuntu and the SSD did show up

wintry yoke
raven forge
humble scaffoldBOT
#

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.

sly sluice
humble mirage
#

I would like to adapt the front-end to

raven forge
sly sluice
raven forge
#

Please share a picture again.

sly sluice
olive ember
#

Is it correct that you provide a partition (/dev/sda1) as a disc ? may be it should be just "/dev/sda" ....

sly sluice
#

It also doesn't work with "/dev/sda"

little hatch
#

Hi folks - copy from #general-archived as I was adviced to sak here instead πŸ™‚

I've got myself in a bit of a pickle. My virtualbox no longer wants to start my HA instance 😦 It says VM "homeassistant" has been successfully started., but its not reachable. Also, if I check vboxmanage list runningvms, there's nothing there. So seemingly it doesn't actually start, and I have no idea why. The host is Ubuntu Server 22.04, running a headless VM for the HA instance.

I would actually prefer to move it to a different machine, so if I could only get a hold of one of the backups inside HA, that would solve everything. But since the VM can't power on, where can I get the backups? The moving of backups outside the system apparently has failed too 🀦

Can't SSH into it, since it's not actually running.. But it's started,

VBoxManage: error: Machine "homeassistant" is not running (currently guru meditation)!
However if I try to start it using vboxmanage startvm homeassistant --type headless, I get...
VBoxManage: error: The machine 'homeassistant' is already locked by a session (or being locked or unlocked)
VBoxManage: error: Details: code VBOX_E_INVALID_OBJECT_STATE (0x80bb0007), component MachineWrap, interface IMachine, callee nsISupports
VBoxManage: error: Context: "LaunchVMProcess(a->session, sessionType.raw(), ComSafeArrayAsInParam(aBstrEnv), progress.asOutParam())" at line 728 of file VBoxManageMisc.cpp

Any brilliant minds with any brilliant ideas I could try out? Or am I simply screwed and have to start over again?

Sidenote: this happened after the last update of Core to 7.2 - seemed like it worked for a day or two, and then seemingly something went wrong there, as it then crashed and wouldn't boot up again. That was the only change to my HA environment for about 2 weeks, running stable for months before that.

sly sluice
olive ember
#

hi @little hatch . If you reboot the server ( ubuntu) what's on the screen of your VM ?

little hatch
olive ember
#

what's the output of your VM when it is booting, and at the end ?

#

you have some kind of prompt ?

raven forge
little hatch
# olive ember you have some kind of prompt ?

I run the command as I've always run it, vboxmanage startvm homeassistant --type headless
The output is

Waiting for VM "homeassistant" to power on...
VM "homeassistant" has been successfully started.
But then vboxmanage list runningvms is nothing, and the VM doesn't actually start anything seemingly.

There's something weird there, because I get some sort of "Guru Meditation" error when trying to run a subsequent command, but I have no idea what that means.

olive ember
#

U cannot run some commands twice ( https://forums.virtualbox.org/viewtopic.php?t=99482 ) that's why i asked you to reboot πŸ™‚
Well, if virtual box is not even able to emulate a BIOS, and if we don't have anything "on screen" somewhere,... then it's a virtualbox failure ...

#

If you can't see what's happening inside your VM, then i have no idea.

sly sluice
raven forge
#

Perfect. I did not expect that to be honest.

olive ember
#

LOL πŸ˜‰

humble scaffoldBOT
#

@runic crag Please do not cross post. Read the channel description, post it and wait for folks to respond. Crossposting wastes people's time as they're unaware of the help you're getting elsewhere.

If you don't get any responses after an hour or more, and your message is no longer on screen, it is fine to re-post or post a link to it.

runic crag
#

Hi i have bticino mh201 lighting automation, i need to install home assistant so i can use it on my phone. For installing it do i need a server? Static ip? Linux? A pc? Where do i start from?

I searched a bit it showed i needed vmware and install home assistant.

Please guide thanks

grand pivot
#

as I said, you should start by reading the website I pointed you to

#

you need to do some research and try things before asking for help

runic crag
#

Thanks, i will

celest quartz
#

had a question about using NUCs and HAOS. When images are being built, there are NUC specific images, but when installing HAOS on a NUC, the x86-64 generic image is used. Should one not use the NUC image for a NUC image or does the x86-64 generic include all those platforms? thanks

raven forge
#

NUC is x86 so that will be fine. Where do you see a NUC specific image?

celest quartz
#

and above it, is the x86-64 generic

raven forge
celest quartz
#

ok, im good with it. NUCs use x86-64 generic

raven forge
celest quartz
#

i should probably do that, with 32GB ram on the nuc, it would have some room

olive ember
celest quartz
#

thats impressive!

#

a NUC?

olive ember
#

something like a nuc .... hp 800-g3 with intel 7500t

celest quartz
#

nice!

olive ember
#

average cpu usage 5-8% ...

raven forge
olive ember
#

No ... I have a cron to drop cache on all VM ... and i use Proxmox in writeback mode.

raven forge
#

But why?

olive ember
#

To release memory back to proxmox.

raven forge
#

I have to try this later.

olive ember
#

@raven forge here is the script:

#!/bin/bash
# maximum memory used by cache in percentage
Treshold=20
MemTotal=$(cat /proc/meminfo | grep "^MemTotal" | awk '{print $2}')
MemAvailable=$(cat /proc/meminfo | grep "^MemAvailable" | awk '{print $2}')
MemFree=$(cat /proc/meminfo | grep "^MemFree" | awk '{print $2}')
Cached=$(cat /proc/meminfo | grep "^Cached" | awk '{print $2}')
compute=$((($Cached * 100)/$MemTotal ))
if [ $compute -ge $Treshold ]; then
        echo "Drop Cache! ( $Cached > $Treshold % of $MemTotal)"
        echo "1" > /proc/sys/vm/drop_caches
else
        echo "$Cached < $Treshold % of $MemTotal"
fi
raven forge
#

That does indeed seem to work. Not sure if it's a good idea though.

spring dust
#

I want to communicate with a device. There is a Python library for doing so, but it gets removed every time I reboot. I have heard conflicting opinions on what might be possible to get the library going. Some said you can’t install anything as a reboot resets everything. Some said to make a custom component, but I would still need the library for the custom component to work. Suggestions?

raven forge
#

More information is required. How are you installing it. What do you do exactly?

empty bloom
#

If your library is custom and not on pypi, then your only options are to add it to pypi or use appdaemon

spring dust
#

Until now, and having never really done much in Unix or Python, I have never heard of pypi. That appears to be a website. β€œThe Python Package Index (PyPI) is a repository of software for the Python programming language.” Repository=Website?

I see tons of posts where others install libraries somehow and do shell commands. Apparently, HA uses something called Docker that you have to additionally fight to install stuff permanently. Even with my library installed, it runs fine from the HA terminal tab, but is not found if I run it from the shell_command addon. I’m guessing that these are two different shell types or it’s a permissions issue.

I see that Unix lets you send ip data directly with the echo and other commands combined. Maybe that’s the way to go?

spring dust
# empty bloom There’s no conflicting opinions. The only way to do this is with a custom compo...

I’ve avoided creating a custom component because I can see from many other posts that it’s extremely complicated and costly. If I could just edit a getting started template, that would be awesome.

However, it seems that there’s the additional rabbit hole of getting it published, since you can’t just put your files somewhere on your system.

It looks like you have to set up a GITHUB account, pay fees, get additional software to post your files somehow, etc.

Again. I’m used to programming in the windows world where I can hammer this out in about 30 seconds, and not have to fight/learn a bunch of different products/services/websites, and have to worry about breaking changes every few weeks. I have at least one HA component that has weekly updates but have not worked for 6 months or more because the authors fix one thing and it breaks another.

empty bloom
#

you can make a custom integration on your system without github at all

#

the integration and files need to reside in config\custom_components\<whatever_you_want_it_to_be>

#

what library are you trying to add to HA? Did you make it or is it made by someone else?

#

Also as an FYI, everything about github and publishing to PyPI is free. So I'm not sure where you're finding your information.

#

paying for github gets you extra perks, but they are not necessary for anything in the process.

spring dust
# empty bloom What are you talking about?

onkyo-eiscp. You can install it with a β€œpip” command, which builds an executable called β€œonkyo”

You can run it from the HA terminal, but the shell_command integration produces:

/bin/sh: /usr/bin/onkyo: not found

empty bloom
#

HA will automatically go out and download & install the version of onkyo-eiscp that's specified in the manifest file.

spring dust
empty bloom
#

It's easier just to use whats on pypi

#

FYI pypi is what builds the index for pip

spring dust
# empty bloom It's easier just to use whats on pypi

This is why I always roll my eyes at modern Unix guys. I just want to write a few lines of code to accomplish something, not learn about pip, pypi, and whatever tools and websites someone made. It seems like a terrible idea for my few lines of code to depend on some else’s project (which could be changed or taken down without warning) which is posted on someone else’s website (which also could be changed or taken down without warning). Am I missing something?

empty bloom
#

you don't need to learn anything about pypi

#

It's like you're being indignant about what you need to actually learn: Making a custom integration.

#

here's the extent of what you need to learn about pypi for the lib you need to add into HA:
Add onkyo-eiscp==1.2.7 to your manifest file.

#

done

spring dust
proven grove
empty bloom
#

it'll still reside in pip

#

unless they remove the code

#

which has happened in the past, but it's rare

proven grove
#

21 years later and there are still python2.7 packages available on pypi.

empty bloom
#

I'm using pypi libs from 2.4

#

granted, they only work on windows xp... but that's a different story

proven grove
#

That... exactly.

spring dust
empty bloom
#

no...

proven grove
#

I still have a version of pyodbc 2.1.8 running on a windows VM from 2010. I feel your pain there.

proven grove
empty bloom
#

Read up on the depedencies section

proven grove
#

I'll put it this way. The largest GIS platform on the planet, ESRI (ArcGIS), uses python for almost everything. Data engineers use it for big data analysis. Banks use it (Ally Bank's entire platform is python.Vanguard Investments as well).

empty bloom
#

all of the best video ai is written in python and is build specifically for windows

spring dust
empty bloom
#

lol, python is not slow

proven grove
#

lol dude... that's a myth

empty bloom
#

Yes it's a layered library, but so is C# which is also widely used

#

Every language that's out there is built on another aside from the ones on the top

#

We use python on our embeded systems here at work to analyze laserdata realtime, which is LIMITED by ethernets slow speed.

#

i.e. ethernet is our bottleneck, not python.

#

and ethernet is really fast.

spring dust
#

C# is compiled and optimized optionally. Python is interpreted, like BASIC, right?

empty bloom
#

so is python

spring dust
empty bloom
#

... yes

#

pyc files are python compiled files

spring dust
empty bloom
#

yep

#

python only compiles when you tell it to, back in the day it used to compile if it detected changes

spring dust
# empty bloom yep

Assuming HA is running interpreted, since you can add Python code to buttons and such

empty bloom
#

HA compiles, python scripts most likely have rules to when it's compiled. It uses a limited subset of python that's locked down. I haven't looked at the code.

#

custom_integrations are compiled at load

spring dust
proven grove
#

lol yeah, if you're looking at Julia or cpython. Python3 uses pypy and is pretty snappy.

empty bloom
#

7th on the list πŸ˜‰

#

that's also prior to 3.12

#

I can't remember what, but 3.11 or 3.12 rewrote the underbelly c files to be faster.

#

3.11 it was

proven grove
#

Yup. Plus, it's not always about speed. Try doing large data sets (millions to billions of rows to process and transform) in C# or Java. There's a reason top data scientists around the world use python over other compiled languages.

#

Yup, 3.11 pushed the execution speed way up. 3.12 made it even better.

empty bloom
#

It will never be as fast as C++, but python is not slow.

#

one of the most annoying myths out there. And the benefits of python drastically outweigh it's speed unless you need speed for your program.

#

and in most cases, if you aren't writing embedded systems, you don't need speed.

proven grove
#

That and the fact that we can write similar methods in way less lines of code. πŸ˜‰

empty bloom
#

Yes, that's what I meant by "benefits outweigh"

#

ease of use is a huge factor to why python is so popular

#

@spring dust anyways, kinda moved off topic here, but if you need help with building a custom integration for yourself, the best place to get help would be #1257019582112334014

spring dust
empty bloom
#

FYI I run about 4 or 5 custom integrations that I've built myself. I do use github as the base repo for my libs

#

I pay nothing

#

it's just used for source control

#

and TBH, you can use anything for source control

#

I used to use mecurial with bitbucket

#

another free source control software

#

and if you don't want to, you don't need to use any source control

#

I used to do that before mecurial

#

But... I do recommend git, once you learn it, it's very powerful and easy to use.

#

it's confusing as hell at first though

amber iron
#

I've been trying to get rid of some stuff from my database that erroneously increased it's size pretty significantly. I tried to purge them with the service and them use the purge service with repack set to yes but my DB is still massive and hasn't decreased at all. Am I doing something wrong? How am I supposed to reduce the db size?

proven grove
amber iron
#

Yeah I was following that. I tried purging it with the repack parameter twice but my DB size has not changed at all and it 100% should be smaller

#

Twice meaning like, once yesterday and one the day before, I know it's supposed to take a while but I have no idea how long or how I'm supposed to tell if it's working

proven grove
#

Interesting... So, no reduction at all? Depending on how large your DB is, it shouldn't take more than a few hours to repack down.

amber iron
#

It's 2.65 GB, something went haywire and I didn't catch it for nearly a week

empty bloom
#

that's not to large

#

what's your recorder settings?

amber iron
#

In the config.yaml?

empty bloom
#

if it's 10 days and you do a purge/repack, it's not really going to reduce the size

#

yes, in the recorder section

amber iron
# 
  purge_keep_days: 7
  exclude:
    entities:
      - light.fan_light_1
      - light.fan_light_2
      - light.fan_light_3
      - sensor.fan_light_1_power
      - sensor.fan_light_2_power
      - sensor.fan_light_3_power```
empty bloom
#

ok, so you're tracking all entities except for 6

amber iron
#

It was 15 before, I just set it to 7 but I figured excluding the entities, which I did recently, should get rid of a lot of it since the dbstats add on said it would be coming from those entities

#

Then i purged with the service and even set days to keep in that service call to 7

empty bloom
#

purge just removes the data, it does not reduce the size

#

repack will reduce the size

amber iron
#

Yes I o had that set to yes

empty bloom
#

but it takes time, a long time

#

HA naturally repacks every second sunday of the month at like 2 or 4 am

amber iron
#

Is there any way to see if the repack is working when I call the service?

empty bloom
#

I don't think there's any feedback, but I could be wrong

amber iron
#

And what happens if HA IS restarted during it

empty bloom
#

I'm not sure, I would assume the process would be interupted

amber iron
#

kay ill try again, see if there was an accidental restart that messed it up or smth

raven forge
#

You can see a temp file being created beside the database file. Check ls -lah /homeassistant/ via SSH. I don't remember the
exact naming scheme.

stray zephyr
#

Hi, I used HA 2024/03 on an QNAP NAS with Container station. I made a backup and wanted to restore the backup at onboarding after the installation of HA 2024/07/30. But, unfortunately, there is no selection to "Restore from Backup" at the onboading screen, as shown in the documentation, There's only "Create My Smart Home". I also tried to restore a backup after I've finished the onboading with the new HA version, but, despite the backups are visible in HA, there's no option to restore any backup. I tried this several times with full removal of HA and all foldes within Qnap Filestation. Every time the same anoying result. What I'm doing wrong?

raven forge
amber iron
lunar siren
#

is it possible to use a USB drive connected to a windows computer as a network storage solution for HA? i have the drive shared on my network

raven forge
severe sinew
#

man i have tried a million different ways

raven forge
severe sinew
#

yeah secure boot is disabled

#

i need to use qemu/kvm unfortunately

raven forge
#

Proxmox VE is. It just has GUI and CLI tools to make it non-painful. I don't remember anyone getting help with libvirt/Virt Manager here.

severe sinew
#

oh man nvm i only had like 3 hours of sleep last night i got it working lmao

#

i was adding it completely wrong πŸ˜„

raven forge
#

Do tell. Maybe it can help someone in the future.

severe sinew
#

im going to take a nap now lmao

sacred island
#

Hi, I am a complete nooby at home assistant. I am creating my HA home for the first time via VM on windows. I am now stuck at the create user page. I have filled in my details and clicked on Create Account. Its been over half an hour with no action at all. Please help?

crude inlet
#

Refresh the page?

sacred island
#

Yes, refrehing did not help

tulip mirage
#

cache clear then a refresh?

sacred island
#

Okaywill try that now

#

No luck

crude inlet
#

In the vm terminal id try a ha host reboot and see what happens

sacred island
#

Rebboting now

#

It appears to have worked, only it doesn't recognise the user name and password I signed up with so I'm unable to login. Thoughts?

#

Forgot Password link refers to methods I can't follow in my browser at the login page

ripe saffron
#

I migrated my Home Assistant from my zimaboard to my unraid VM. I cant get a local connection from HA.local. I have access to everything else: nabu casa and ip_address:8123. Just not local. It still uses my old HA for local. How do i move it to the new Unraid VM for local access.

raven forge
#

Research unraid mDNS. I'm not familiar with unraid but it might prevent that in some way.
I don't recommend .local anyway as it's finicky and unreliable in my experience. Just use hostnameofdhcpclient on its own. Most good routers support that if you use their DNS server.

ripe saffron
tulip mirage
raven forge
ripe saffron
#

right now i can connect to ipaddress:8123 just not Ha.local:8123

raven forge
#

I understood that. Have you tried with just the hostname without .local?

#

If you want to fix the .local thing we need to know more about your network and your VM's network config. I'm just saying that I wouldn't use it. Use something that reliably works.

#

Oh wait a second. You have two HA's running and homeassistant.local refers to the old one. Did I get that right?

ripe saffron
sacred island
raven forge
#

Ah. Sorry about that. Check the name here: https://my.home-assistant.io/redirect/network/
And try what ping homeassistant.local or whatever the name of the instance is responds with.
Does the VM use bridged networking and is in the same subnet as your PC?

ripe saffron
raven forge
#

I want you to check the host name of your VM's HA instance and answer the other questions πŸ™‚

ripe saffron
#

This is my VM setup

raven forge
#

Looks like a bridge. That should be okay.

#

Ah. It says homeking.local.

#

Try to ping that.

ripe saffron
#

omg that is it lol

#

that worked lol

#

Thank you @raven forge

raven forge
#

No problem. Sorry for not properly reading the first message.

ripe saffron
ripe saffron
raven forge
#

Zwave addon? Remember that addons are in their own container hence localhost refers to itself.

ripe saffron
#

ok ill look at the logs to see the error

soft shadow
#

Does anybody have experience with setting up Proxmox for HA? I had a Pi which stopped connecting to the internet (no wifi interface anymore) and bought a cheap NUC, and now I am wondering if it could be a network issue (aka I think devices might be wandering off into different subnets on their own) - wondering if anybody who enough to help out (ideally with the NUC, since its more powerful than my own 3b Pi even if its just a network config issue).

stray zephyr
raven forge
#

Only HAOS has restore functionality. You have to restore the files manually.

olive ember
#

Hi. I'm fighting with zigbee2mqtt. Could someone give me the firmware i have to flash on my sonoff zigbee E version ? i'm bocked at the error "zh:ember: Delivery of BROADCAST failed for '65533' ". I don't find anything except rollback to previous version (zigbee2mqtt and or FW and/or ember->zqm)... I'm lost.

raven forge
olive ember
#

Thx @raven forge πŸ™‚ There was a sign "Beware of the trap" ... guess what ;-)))) explode
... "If you are not using the Mosquitto broker addon fill in your MQTT details" LOL

hexed maple
#

I'll also ask here:.. I think I'm hitting the hardware limits of my raspberry pi4.. how big a deal is it to move the config over to a docker/vm setup instead?

raven forge
#

Depends on your background. I consider it very easy. For docker it's basically just moving some data in the right place. For HAOS/VM you can use its restore functionality.

stray zephyr
grand pivot
#

you just extract the backup into your config directory

stray zephyr
chrome locust
#

hello all! when installing HA directly thru the image for Rasbperry Pi, can I still ssh to raspberry? I'd need to ssh into it to install HACS

tender wren
#

No (if you did not install one of the SSH add-ons)

#

Use the install instructions for OS/Supervised