#installation-archived
1 messages · Page 68 of 1
The CCNA basics class is in session
next you're going to be ordering the octopus
4 bytes, each of 0-256 in size
minus the multicast
or wait, the multicast counts too
it's just 0 through 255
which is 256 ion total
128, 64, 32, 16, 8, 4, 2, 1,
0 0 0 0 0 1 1 1 = 7
1 1 1 1 1 1 1 1 = max 255
0 0 0 0 0 0 0 0 = min 0
and 254 of them are usable
minus the multicast indeed
and the gateway
next we'll be playing around with ip fragmentation and mtu sizes
no we will not.
ok
sorry I'll stop
I gotta figure out the mac of my router to see if I can make a ARP poison attack rule in wireshark to figure out if anything is getting poisoned
thanks for the help everyone
Hey im cofused to how to.
i have here a rasberry pi 4 with broken usb ports and a cb1 (less powerful rasberry clone)
I would like to use the rasberry as my main home assistant device which is already installed and running.)
What can i install / how do i install the cb1 as a local screen in my personal room and maiby as a file acces (i would like to add the external ssd to the cb1)
you could use some sort of network storage like NFS or SMB perhaps
and a screen shouldn't be hard, you just need a full screen web browser and the address of your home assistant. You could even run it in some kind of kiosk mode depending on the OS you install
a touch screen is even cooler
I think I'll be using some sort of tablet for mine
it is a touch screen with the cb1 (not with the rasberry)
tft screen as far as i know
cool
rn i have only found this precompiled in the guiithub page
i dont know how to really make screen display the webpage to be honest
CB1_Debian11_minimal_kernel5.16_20240319.img.xz
I'm not sure what the cb1 is exactly, but I figure if it runs linux, you could probably get an X session on it to display a window I guess?
X11 or wayland
xorg
there's loads of window managers and desktop environments to use in that case
hmm okay thanks for the info so basecall for storage that i can connect to home assistant i can use;
- NfS
- SMB
For the screen i could try:
- X11
- Wayland
- xorg
Thats gonna be fun to try
use what you fancy
it's not easy, I can tell you that
for linux you'll want to choose a desktop environment to install, which is easier then manually configuring xorg/X11 or wayland
there are tons of them out there
@shy flare Just so I understand you correctly, you want to run HA on a rPi and have a screen attached to that as well displaying the dashboard, correct?
like Gnome, KDE, cinnamon, enlightenment, pantheon, mate, and a bunch of others
LXDE, XFCE the list goes on and on
no, he wants to have the screen attached to the cb1
which is running a debian linux distro if I'm not mistaken
More or less i already have Home assitant running on a rasberry thats more powerfiul
The cb1 have a screen and it cannot connect to the rasberry
main priority is to get the screen (touchscreen) to display the dashboard (that have controls)
and the home assistant is on the raspberry pi
well, if you get a bit famillair with the TTY, it is definitly possible
if the cb1 has little resources, I suggest a lightweight desktop environment
LXDE or XFCE are probably good choices
you'll need a display manager as well, to start the desktop environment when you log in
Whew... ok. Yeah, that's actually pretty easy to do. Honestly, the drop dead easiest way to accomplish that is to just run raspbian desktop and just put Chromium into headless kiosk mode. There's a lot of tutorials out there to accomplish that.
it's like a login manager
the thing is though, the raspberry is not the one with the screen, so it's not gonna be running raspbian
either way he'll have to set something up on the cb1
The CB1 is a rPi clone. It should be able to run raspbian.
once he has a desktop environment and a login manager on his debian linux system that he was about to install on the cb1, then he'll be able to start a desktop and open up something like chromium (a open source fork of the chrome web browser)
well that is pretty easy then
get a distro with a pre-configured desktop environment like raspbian
and just open up chromium
and set it to kiosk mode
if it can run debian, it can probably run whatever flavor you throw at it anyways
but I like raspbian, it's pretty lean for small resource devices
Okay
basecally installing rasbian on sd card setup chromium and a kiosk up
much to learn but i get into that thank you guys very much
that should be the easiest route, you can also delve into a TTY with debian minimal, which is more hardcore 😛
we can make it a lot more complicated too if we put gentoo into the mix and having to compile your own linux kernel
or maybe linux from scratch, but I don't really recommend that unless you want to take weeks to get a project finished
but it is a nice learning experience if you go that route lol
Like i could do that i did some other compiling in the past but linux kernel might be a few levels over my skill 
I went through it a couple of times, there is a LOT to configure, I can tell you that
KISS principal (Keep it simple [redacted] lol)
it would basically mean having to configure which features you want your linux kernel to support for the specific hardware that you have, you'd have to be pretty well versed in the hardware that you are using to figure out which type of kernel level features the cb1 needs
so I think it's not a good idea
I'm just mentioning it because it could be a fun side project when you have a lot of spare time
yayit works like a typical apple product.
Well as a small update i figured out that the cb1 only use their specific firmware
any other firmware dont work.
back to the hard way :)
Hi all,
running HA on my Pi4 for quite a good time, the Installation is on an SSD attached via USB.
Today i would have liked to update HA OS from 12.3 to 12.4, but the Update fails with the MSG Popup:
Failed to call service update/install. Error updating Home Assistant Operating System: Unknown error, see supervisor
When looking through the Supervisor logs i found the following entry:
2024-06-20 16:08:54.014 ERROR (MainThread) [supervisor.os.manager] Home Assistant Operating System update failed with: Installation error: Failed updating slot boot.0: failed to run slot hook: Child process exited with code 1
any advise how to fix this? Searching with google ony ends me in topics without solutions (beside one that downloaded Backup and startet over with fresh install) but until now i would like to avoid this if possible
Most likely you have two HAOS installations on different media on that Pi. Did you install HAOS on the SD card as well as on the USB SSD and have both connected? 🤔
Hey, does the companion app need any specific port setup?
ive put my HA behind a nginx proxy and the companion app refuses to make connection eventho the same domainlink nearly instantly loads the page in my phones browser
Which OS? #ios_and_mac-archived or #android-archived might know better.
Hello, I just picked up a RPi5 and 128GB SD card, used RPi Imager to flash Home Assistant. I put the SD card in and have it shown up on HDMI to watch it boot, it just stays at Bootloader and looks like it doesn't find the partition, and it just keeps looping trying everything. Also, the NIC isn't activating....hardware issue?
You did flash the image for the 5, and not the 4 or the 3?
correct
i specifically chose Rpi5
odd that the NIC won't even activate, let me try a new cable real quick
known good cable on a known good switch port...nothing on rpi5
try reflashing the card
did twice already
what brand is the card?
ah, ok. you didn't say canakit so I thought maybe it was a fake card
have you tried flashing a different OS to it?
nope not yet....guess thhat is next step?
it would certainly be a step I'd take
trying now, will report back
yup, RPi OS works just fine and network works too
so its something with Home Assistant
i thought the latest release for Rpi5 was stable
I am trying the RPi5 Home Assistant OS flash one more time.....
Nope.....says stuff like this:
Trying Partition: 0
Trying Partition: 63
Failed to open partition 63
Trying Partitin: 1
Failed to open partition 1
Boot Mode: USB-MSD (04) order f
Any ideas?
Flash an older version of HAOS?
I wish I could, but the Raspberry Pi Imager only lets me choose the most recent option
On android
Turns out it was setting related, I could login but after the authenticator it would die everywhere, didn't notice untill I logged out on my pc
Seems like other people are having an issue:
https://forums.raspberrypi.com/viewtopic.php?t=372554
Yup, officially an Issue:
https://github.com/home-assistant/operating-system/issues/3437
I'm 3-5 hours into importing a partial backup (75MB from a working python venv/core install--that's now been wiped) to a fresh haos install. Cli "jobs info" shows the backup saying "done: true" and "progress: 0" and "stage: docker_config" Is that typical? I know backup imports can take a long time, but I can't tell if its hung or not.
Reviewing the logs, the server starts the backup restore, and then immediately (within 15s) says its complete. The job remains listed in "jobs info" and I've also tried restoring the backup after starting onboarding to the same log results. My server isn't huge, such that manually rebuilding it wouldn't be impossible, and it's starting to take as longs as rebuilding would require. Am I better off calling it a loss and just manually rebuilding it?
No no sd card in the pi. The system was freshly installed onto the ssd and the pi has only seen a SD card once to update to be used as ssd boot.
ok, today i have again restarted the pi and booted ha in safe mode, after then restarting it once again it was working.
Greetings! When I get "Home Assistant CLI not starting! Jump into emergency console..." is there a set of files I can review to figure out what went wrong?
What hardware are you running HA on? Do you have an active internet connection on that device?
That was it. Bad network cable.
running journalctl told me that it couldn't reach DNS
Thank you!
i'm not sure where to put this. I have a intel mac mini from 2012, i have a 128 GB HD and 16gb of Ram. What would be the best way to install Home Assistant on it. I don't need mac os on it or anything, could run on ubuntu or something else on the SSD whatever is easiest
You could throw Ubuntu on it with proxmox and run HAOS in a VM. You're not going to have the best experience, but it'll work.
what would be a better experience? i'm trying to not have to buy new hardware if i can avoid it.
is the docker route no longer recomended ?
Well, on that old of a Mac, you're gonna run into I/O contention. But, it won't be too bad.
Oh, yeah, you can go with Container, absolutely. But most people want addons, so first suggestion is always HAOS.
I recommend that people use Container if they know/love/want to know Docker, otherwise HAOS
Or, if they hate themselves, they go Supervised
You spoke the name of that which we don't speak the name
So to do haos what is the best option then? This Mac mini is just sitting here doing nothing
it's 12 years old, so that's not surprising
It depends on if you need/want addons or if you are ok with docker AND don't need/want addons in HA. Personally, I use Container myself as I don't need or want addons and I know docker really well.
I don’t know docker well. I’m not even sure if I want add ons yet I have not done anything with home assistant before.
In that case, I'd go with either installing HAOS directly on the Mac OR going with something like Proxmox. The first option is the easiest by far and since you're just starting out, that's what I would recommend.
but remember that it's a 12 year old computer and that those aren't very good.
But it runs a pi4? How much better is that than that Mac mini?
Apple dropped support for that system in 2020.
looks like a 3rd generation probably i5 cpu
The 2012 Mac Mini has lots of issues with I/O contention (the onboard bus SUCKS). Plus, CPU contraints and such. You can throw a SSD in it and that will help. But, generally speaking, the chipset on a rpi4 will beat a 2012 Mac Mini computaionally.
ARM processors are RISC, and so things that are written for the platform are a bunch more efficient. also, though it doesn't matter much for HAOS, the pi4 has hardware encoding/decoding for various media types built in and intel didn't add that for 5 more cpu generations
Ok I also have a m2 Mac mini that is always on. But I can’t run HA ont hat right?
Sure you can. But, in that case, I'd definietly go the VM route.
I don't think you'd want to dedicate a m2 Mac to HA. Kinda overkill.
It’s my everyday one. Has scrypted on it and some other things.
Basically my late 2012 too crappy. M2 too nice
Eh, if it's what you have, give the 2012 a shot. I have one here that runs some basic, nonessential things and while it struggles from time to time, it still works.
When you all say add ons what does that mean? Like. Adapters for zigbee z wave etc ?
addon is a container in HAOS,
for things like NUT that I think run in docker? newb here too lol
Not exactly, but also yes. lol So, an addon is basically applications and functionality that run within HA itself (as docker containers).
Basically (for both of you lol) addons do run in containers, but within HA itself. They have some special "magic" that enables them to communicate via the HA UI.
Yup, that would be either a Core install or Container.
Supervisor is the word I was reaching for
We don't talk about Supervised here. 😉 It is the most difficult of the installation methods and the hardest to support.
LOL, I only know of it because I'm running RaspberryPi HAOS and had the dreaded "system not healthy thing" and while researching found the snipit that you can't run add-ons on Pi HAOS without Supervisor (whatever that is).
(Not supervisor, that's different. Sorry for any confustion).
Thank goodness the full backup and recovery worked 100%. 5 stars, will implement again.
Ah, yeah. That's correct.
Yup. That's it basically.
I’m still confused 🫤
lol ask questions and I'll try to resolve your confusion.
(Wow... that sounded way too AI-ish)
an add-on is a separate program that HA can run in a "VM" like thingy along side the HA program itself.
Like runing OBS to capture a game stream, kinda (but not really)
I just wanted to play around with HA for automations from smart home stuff
Correct. So, for instance, take an app like Plex. You can "run" a Plex instance within HA. It runs as a docker container on the same hardware as HAOS is running.
To answer your orignal question. No. Zigbee or Zwave are devices, not add-ons
Add-ons are apps that give you extra functionality
You can absolutely do that. You don't have to run addons, but you have the choice to if there is extra functionality that you might want.
Zigbee / Zwave goes in devices, like Roomba or Roku etc...
Weeeeeelllll that's kinda correct. You can run Zigbee2MQTT, MQTT and ZWave2JS as addons within HA.
Ouch! lol
I used the newb auto detect so they are devices for me
You don't HAVE to, but you can. I run those as separate docker containers on another machine.
But some people prefer to have everything in a single UI/single instance.
my only add-ons are Duck DNS / Glances / NUT / Node-RED / SSH
Ahhh!
I'm stuck on getting my NAS to com to my NUT.
Anywhere else I'd be banned for saying that lol
I saw the question, but I can't help with that 😦 My APS is connected to my NAS, but separated from HA.
I saw a comment about a wireshark alike for sniffing packets in HA?
Otherwise I'll have to mirror a port and see if there is anything obvious
The NUT guys don't have discord, just IRC and listserv lol! Any excuse to update my nicserv ID I guess...
And I guess I posted in wrong chan, that should have gone in add-ons I guess. I'll cross post. Better to ask for forgiveness later as they say.
I cant find the configuration.yaml on my HA
So what’s the easiest way to go about that. Haos? On the macmini? Docker on Ubuntu?
Personally, I'd go with installing HAOS on the Mini and see how it works for you. Docker on Ubuntu is fairly simple as well (the docs are good on the install there). BUT, since you're a new user, I'd go with HAOS and see how you like it first.
Are you still running Container or did you switch over to HAOS?
currently running just HAOS i believe
Just use the file editor then. 🙂 Simple solution. If you really want to find the actual file, use the Terminal. It's right in the root of the directory.
Hey guys, im running the latest HAOS on hyper-v. however, within the last month or so, i've been experiencing Zigbee issues. i've narrowed this down to weird inconsistent network issues. if i ping one of my network devices from the VM i will sometimes get a reply and sometimes for a short period it will timeout.
Anyone have any ideas? i've disabled VM Queues for the network adaptor, still getting inconsistent results
Is there specific one I can follow to do this?
Is the file editor on HAOS on the actual Pi or is it on the web page for HA?
On the webpage for HA.
where.... I couldnt find it
https://www.home-assistant.io/docs/configuration/ (I think you have to install the file editor addon).
Is an online machine required?
if you're asking if you need Internet access to install HA, yes
no I'm talking about having a PC that remains open all the time to leave the site open
Yes. The machine that runs HA needs to be running.
okkkk
Ok I was looking at the Linux stuff and it was saying to do docker and all the stuff we had said not to run haha
lol you could do that, but I'd still say try it this way first. I think you'll get a better experience.
you do not need a seperate PC on with the webpage open if that is what you are asking
No I want to do that. Just trying to figure it all out. Mac mini doesn’t have bios stuff like those machines
Oh, yeah, I probably should have mentioned that you can skip the bios stuff.
So basically if I’m currently running Ubuntu on a SSd that is attached externally. I can just download haos? And install it on the SSD drive I have inside the mini?
🤔 Yeah, I think that should work
My stuff is all backed up from the internal drive (old Mac Catalina stuff) and I moved it all to m2 and did backup there too
So by running that installer from Ubuntu onto a new SSD it should boot into it?
I guess worse case scenario if it all goes to crap just run docker on Ubuntu that currently boots haha
Yeah, that should work. Or, if you have a thumbdrive, I think you can use that as well.
I guess next step is put the new SSD in.
can someone help me with setting up lightwave for the first time on HA, this is what i have done so far lightwave:
host: 192.168.1.xxx
lights:
name: Archie Light
switches:
name: Archie's Blind
Gah. ignore... IP Conflict on the network...
I’m not sure I can get this to work. Really was trying to not buy a new machine to do all of this.
Or just not smart enough to figure it out haha
I’m mobile right now, but I can jump on tomorrow and give you a hand.
Nah not your job for that stuff. I just don’t think I’ll be able to use this machine which is frustrating
hi y'all, I have migrated my long-time docker homeassistant installation to a new rapsberry pi, I have a backup of the whole /config including .storage, but when I start the docker image on the new pi with the same config folder mounted (same compose file), I'm greeted with onboarding screen. I have went through the onboarding and I can see all my devices are still there as well as my dashboards and also users, but I can't seem to log in using the old logins. how can I properly restore this docker setup?
from the container I've tried running hass --script auth --config /config list and it lists just the one user I've created during the onboarding, but when I navigate to People, I see all the previous users listed
Please share your compose file and perhaps the output of running something like find /pathtoconfig/ ! -path '*www*' ! -path '*custom_components*' ! -path '*blueprints*'.
ok so I figured it out, turns out not all the files from /config/.storage got copied, namely auth and core.config and some others were missing among the files I've moved over, turns out they weren't copied due to permissions, so I've readded them and it's back up and running. phew. learnt something new about backups.
I can recommend rsnapshot for simple file backups. Very easy to use, explore and restore as it's just files/directories and hardlinks: https://rsnapshot.org/
Since HA involves a database I'd also recommend doing backups with its own mechanism so it doesn't corrupt: https://github.com/jcwillox/hass-auto-backup
I suppose you can also automate it with a automation: https://www.home-assistant.io/integrations/backup/#services
ooh thanks, I'll look into that
Hi all. I have been running HAOS inside a VM on proxmox for 2 years now. As I am reluctant to use sqlite for huge/critical DB, i created a VM with postgres one week later. It has been "rock solid"!!!. And in case i have a "glitch" with a new version, i can rollback the VM running HA without impacting the DB. Great!!! I am now realizing that i would like to tweek some stuffs here and there ( swapfile, memory, cron, ...). I guess the best choice is to migrate to "supervised" installation. Any recommendations ? Can I use the backup/restore of HA ? will it keep all my device id, add-ons, configurations ? thx.
Don't use supervised:
Home Assistant Supervised is considered advanced and should only be used if you are an expert in managing a Linux operating system, Docker and networking. Maintaining a Debian installation to a very specific set of requirements is hard.
Insecure installations and Unsupported configurations will result if you dont know what you are doing.
You MAY regret using this method.
If you want an appliance use HAOS, if you want control use the docker installation method.
Why do you feel you need to tweak things? Why do you want to change swap? What do you mean with tweaking memory? What do you want to run via cron?
i got an error installing home assistant
i can’t read the log file because i don’t know how to. anyone able to help?
ha supervisor logs and ha core logs via CLI.
sorry i dont know what that means
You need to enter this via keyboard and monitor on the command line.
We can't really troubleshoot much then. Can you share what you saw at least?
Sure.
it seems the file doesn’t want to attach
ill just paste a bit of it if that’s fine
it would appear i cannot attach any files here
allg if i send u a dm instead?
Please use a code share site to share code or logs, for example:
- https://dpaste.org/ (select YAML for the language, and consider picking a longer expiry)
- http://pastie.org/ (select YAML for the language)
- https://paste.debian.net/ (you guessed it, select YAML as the language)
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.
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.
ill do what the bot said 😭
This 500 error can have a few reasons. Often times it's DNS related.
Could not resolve host also hints towards that. I'd have you run ha host logs had you the possibility to.
is it possible to access the pi remotely
Not in this state. Also check: https://www.home-assistant.io/more-info/unsupported/dns_server/
You might be able to enable developer SSH but I wouldn't recommend it if the word CLI doesn't mean anything to you: https://developers.home-assistant.io/docs/operating-system/debugging/#enabling-ssh-access-to-the-host
Where are you installing HA on? How is it connected to the network and internet?
raspberry pi 4b
connected through an ethernet connection directly into my laptop, which is subsequently connected to the internet via a mobile hotspot
wait i restarted the pi and there's a new error log
Always the weird network setups 😄
dont even question it 🙏 i got one mod already laughing their ass of at my setup
I'm absolutely not familiar with that kind of setup but I'm not gonna laugh at it. Hmm, I just noticed that I contradict myself with the emoji but I was laughing at the situation.
I'm not really sure how the laptop is supposed to give network via the ethernet port to other devices.
I don't recommend it and am not familiar with this either but perhaps you can let the pi connect to the hotspot as well? https://github.com/home-assistant/operating-system/blob/dev/Documentation/network.md
Maybe but trying to debug the network via SSH isn't necessarily a good option in this case. I linked to the official docs how to enable SSH. What kind of hotspot is it?
its a uh
iphone hotspot?
i ytierd running a ssh pi@192.168.2.2
but it refused to connect
i manually configured the ipv4 btw
Oh. So Iphone > Laptop > Ethernet > Pi > HA. So your HA relies on WiFi, mobile data, a laptop and WiFi to ethernet bridging 😬
There's so much that can go wrong, so many variables, so many dependencies and the complexity is quite high and the reliability low. What if you leave the home, have to update the laptop or want to take it somewhere else?
just wait till you hear about the $10 router that i will connect the pi to once im done onboarding
with no internet access btw
If you only have mobile data I'd recommend a LTE/4G/5G router with ethernet and a switch.
Note that developer SSH runs on port 22222, not the default 22. There's also no pi user.
turns out there was no ssh file on the sd in the first place
so i did a quick sudo touch ssh somehting
its still refusing connection though
Pretty sure HAOS has no sudo either. Unless you mean you ran this on the laptop, of course which makes more sense now that I think about it.
Unfortunately your setup is wildly unfamiliar to me. I don't own any apple products either. Do you only have mobile data? No "real" internet?
unfortunately yes
ill keep tinkering a bit
would ya fucking look at that
the create homeassistant button is here
all this fucking around managed to make it work somehow 😭
THE DASHBOARD
IT'S BEAUTIFUL
FYI, I restored the "full backup" of HAOS into HA Supervised (I used Debian 12) ... everything is working! history, zigbee, tuya, esphome.... and a keyboard in FR, that's a game changer!!!! 😉 😉
native ssh with copy/paste....
Let's just hope it stays healthy: #installation-archived message
Debian is great though: #installation-archived message
Supervised, the thing that should not be 🤣
Hi,
I need help setting up Home Assistant on my server.
Im currenty using VMware and i followed all the installation steps.
The last step is what i cant find in my folder:
"Right-click the .vmx file and select Open with > Notepad.
Under .encoding, add a line. Enter firmware = "efi"."
Hoping someone can help me with this, Thanks 🙂
This line was taking out of the logs,
2024-06-22T11:43:09.017Z In(05) vmx DictionaryLoad: Cannot open file "C:\Users\nadav\AppData\Roaming\VMware\config.ini": The system cannot find the file specified.
HAOS is like iphone .... we need to jailbrake it to have full power 😉 😉
lol no you don’t.
Sorry, i'm using proxmox, but indeed, HAOS is difficult to install 😉 ... because of qcow/uefi !!!! ( i completly forgot that)
That’s still 100% better than Supervised.
Try to install it bare metal on x86 😄
my HAOS vm is stopped .... I don't see what could go wrong, as everything is docker.... 😄
can someone help me with setting up lightwave for the first time on HA, this is what i have done so far lightwave:
host: 192.168.1.xxx
lights:
name: Archie Light
switches:
name: Archie's Blind
I'm (attempting to be) a new HA user. I'm trying to install HAOS on an ASUS NUC PRO. I tried the Generic x86-64 install method (https://www.home-assistant.io/installation/generic-x86-64#start-up-your-generic-x86-64), as well as pulling the NVME out and writing the HAOS image (v12.3) directly to it. Both result in it stalling with the message "Booting `Slot A (OK=0 TRY=0)'". Any idea what I could be doing wrong?
did you do the part about secure boot and uefi?
yeah. Bios is config'd as instructed
do you have anything plugged in to usb?
a Kb & Mouse..
not really necessary but I guess not a problem either
& had the USB mem stick when trying the install from Ubuntu of course
sorry to say that I don't have other ideas, except maybe verify the bios settings and maybe toggle them and reboot and toggle back and reboot
but that's just from searching your error
I'll give that a shot.. thanks.
Hi, Im setting up an haos system for a girlfriend of mine. I got a pi4 4gig. im migrating from docker to kvm. but i get this error.
[supervisor.plugins.dns] Detected a DNS loop in local Network!
it runs on bridge mode.
oh. I have the bridge network br0 used and not the kvm bridged network
Hi,
I am trying to reinstall HAOS on odroid N2+. The server starts and I am able to ping it. But it gets stuck at the setup page and gives errors that it cannot clone the add-ons repository. Anybody facing this issue?
Do you have internet access on the odroid?
I installed home assistant via docker but when I want to connect to HA, I can only from the PC which serves as a server but not from other PCs on the network
Please share your compose file and address you use to connect to it.
what is the compose file ? and the address is http://172.17.0.1:8123
you mentioned that you installed via docker, do you have the lines you used to start the instance?
Something I was hoping you use: https://www.home-assistant.io/installation/linux#docker-compose
The docs should really list that first and recommend it.
Where did you get that address from? Try the ip of the host running the container instead.
--name homeassistant \
--privileged \
--restart=unless-stopped \
-e TZ=MY_TIME_ZONE \
-v /PATH_TO_YOUR_CONFIG:/config \
-v /run/dbus:/run/dbus:ro \
--network=host \
ghcr.io/home-assistant/home-assistant:stable```
if you see there there's no port mapping, you're missing -p 8123:8123
MY_TIME_ZONE and PATH_TO_YOUR_CONFIG are place holders.
HA uses host networking instead of port mappings so using -p would be wrong.
first time I hear about making containers use the host's networking in my life. The issue is somehow the networking config in the host, the 172...1 ip is normally a router/gateway, not a normal dhcp given ip address. Can you give more details about the host?
HA requires host networking in docker. Been that way for at least a few years now.
Docker uses 172.:
- https://docs.docker.com/engine/tutorials/networkingcontainers/
- https://docs.docker.com/network/network-tutorial-standalone/
Not sure where they got the ip from exactly as they didn't answer my question yet: #installation-archived message
the host is a PC connected to wifi
my bad i used the bad ip 🥲 sry
🤷
Its connected to the same router for the past 3 years and has run without any issues. The SD card failed and I had to reinstall. So I assume it has internet access.
I installed HAOS on RPI3 without any issues. Looks like an issue with odroid N2+ image. the log is attached below https://dpaste.org/YCfqv
I posted in General but I think this might be a better place.... I am trying to change the radius of my home zone from 100 to 200 but not sure where to do it at.
@grand pivot Thanks. So I have read that before but allowed myself to get bogged down. Just read it slowly and tried adding to the configuration.yaml file and I see it worked.
Hi. I'm getting low disk space warning on my 64GB HAOS VM. Deleted some backups and I'm at 82% usage. I have run this, but can't find what's taking the space. For example, DB file is not listed here:
du -h / | sort -rh | head -n 20
may be you could try : find / -mount -size +20000k
thanks. homeassistant folder is 14GB, backup is 10GB. Everything else is 1GB. So I'm at 25GB out of 64GB. This does not add up :\
Which option did you follow? You have to use the system level way to see all.
Well, I don't know. On my HAOS system, i could not find my GB also....
weird.
yes... and trying to find/solve using the wrong keyboard got me crazy
for sure your /des/sda8 is full ( nearly ) ..
If u are using esphome, you could try to "clean build file" on each device.... It can't hurt.
Having trouble installing HAOS as a VM on my unraid server. Added the qcow image and booted, setup is having issues:
2024-06-24 04:26:46.592 INFO (MainThread) [supervisor.docker.interface] Updating image ghcr.io/home-assistant/qemux86-64-homeassistant:landingpage to ghcr.io/home-assistant/qemux86-64-homeassistant:2024.6.4
2024-06-24 04:26:46.592 INFO (MainThread) [supervisor.docker.interface] Downloading docker image ghcr.io/home-assistant/qemux86-64-homeassistant with tag 2024.6.4.
2024-06-24 04:32:21.992 ERROR (MainThread) [supervisor.docker.interface] Can't install ghcr.io/home-assistant/qemux86-64-homeassistant:2024.6.4: 500 Server Error for http+docker://localhost/v1.45/images/create?tag=2024.6.4&fromImage=ghcr.io%2Fhome-assistant%2Fqemux86-64-homeassistant&platform=linux%2Famd64: Internal Server Error ("Get "https://ghcr.io/v2/": context deadline exceeded")
2024-06-24 04:32:21.992 WARNING (MainThread) [supervisor.homeassistant.core] Error on Home Assistant installation. Retrying in 30sec
posted this question on forums, hopefullt someone will have an idea.
https://community.home-assistant.io/t/cant-figure-out-why-im-running-out-of-storage/742776
No Sorry...
Huh?
Just to check, the VM has working internet access?
I have no idea how to check as theres no terminal, vnc takes me to the ha "terminal" with only a few commands
VM has the same network settings as my other VMs which have internet
Not sure whats up with the IP in the network info, as that's not actually it
That is the container IP address.
Is HAOS just a linux distro running docker version of HA?
But how can I view the network settings for my actual connection?
Yeah, basically. HAOS is a light Linux OS and then everything inside of it runs as containers.
What are you using for your VM controller?
unraid
If thats the case, why is there stuff exclusive to HAOS and not in the docker version? Sorry if dumb question, very new
In your VM settings, is your network souce set to your main network?
Because they are different things. The OS is able to manage the containers and itself. Addons are containers.
Because the OS controls the containers.
So I could in theory just host addons in docker manually?
And not use the OS?
Well, sort of
Yes, identical to my other VMs
Add-ons mostly only work with HAOS, but add-ons are almost always based on existing containers, or there's a container option from somebody else
Deleted the vm/qcow file and restarted, I think it might be working now
Ah no, just took longer than I thought to get to the error
2024-06-24 06:38:32.236 INFO (MainThread) [supervisor.updater] Fetching update data from https://version.home-assistant.io/stable.json
This is indicative of it connecting to the net, isnt it? Or does it just not have a log for failing that
Also error 500 on the pull for the container indicates server issue doesnt it?
@vast sapphire generalizing to all HTTP requests, 5xx is supposed to indicated a server issue, but some will incorrectly return that if an invalid request causes an error, for example
What's the image tag and repo you're trying to pull?
Oh, it's a HAOS update
Just initial install
I havent modified anything, just loaded up the image in a vm, and it started installing. Then gives error:
2024-06-24 06:44:56.965 ERROR (MainThread) [supervisor.docker.interface] Can't install ghcr.io/home-assistant/qemux86-64-homeassistant:2024.6.4: 500 Server Error for http+docker://localhost/v1.45/images/create?tag=2024.6.4&fromImage=ghcr.io%2Fhome-assistant%2Fqemux86-64-homeassistant&platform=linux%2Famd64: Internal Server Error ("Get "https://ghcr.io/v2/": net/http: request canceled while waiting for connection (Client.Timeout exceeded while awaiting headers)")
Network issue on your end based on that last bit
How is it able to fetch update data earlier in the install though? DNS maybe?
Could be any number of things (including DNS). It could also be a transient issue, but I assume you've retried it a few times already?
Yeah
this is the full log. Lots of stuff to indicate its connecting out
But I could be wrong
Does the issue occur consistently? Or is this your first attempt?
Yeah have rebooted 6 times, deleted VM and started over once
OK people, I need a solution for my setup as I'm rethinking the structure
I have a mini pc (elitedesk 800 g2) running proxmox, within that I have 2 vms.
- windows running torrent and htpc (storing attached usb drives)
- hassos
Now I'm incorporating a Frigate setup and would like to install it completely off hassos, an independent installation.
So I'm thinking to install it in a docker on proxmox. Problem is, the USB drives that are attached to windows would also need to be attached to the frigate docker which isn't possible. Any ideas?
On the other hand, I was thinking of completely redoing the setup...
Put Ubuntu on the mini pc (no proxmox). Then running everything I need on seperate docker containers - windows, hassos, and frigate within that Ubuntu host
Possible? What are your thoughts?
As a disclaimer, I'm running HA as a docker container, rather than HAOS. If I was experiencing this issue though, I'd SSH into the host and open a shell on the HA container to ensure that I could resolve the relevant DNS names and curl relevant URLs.
The GitHub returning 500 error codes for artifacts is a pretty common issue with them
Try again in a couple of hours
It's not a container, it's a VM. Don't i need to enable ssh which requires finishing the install?
Vnc just gives me access to the ha "terminal"
Oh i think i misread what you meant my mistake
As I understand it, HAOS runs HA and add-ons as docker containers internally. I definitely could be wrong about that though--haven't read much into that section of the docs. It's also a bit odd that it's trying to pull images if it's not implemented in that way.
If so, you can use docker exec from inside the qemu guest to run the necessary commands to install/start sshd/dropbear and add your public key.
Well, yes, but no, and that's terrible advice
HAOS does use Docker - see the image pinned in #general-archived
Spinning up containers with the docker command can, and will, break stuff
docker exec runs commands on existing containers, by the way
There's no need, or point, in installing sshd inside a container
Everything you need to do can be done via the console
How can I ping google.com to know once and for all if theres internet?
login to get past the ha prompt
Whats the login?
ah
(well, hit enterr)
That was the answer I was initially looking for lol
yeah it definitely can reach the web
so the issue is likely githubs end
It used to tell you that on the console, but they changed it
Will HAOS ever have a simple offline iso installer?
Check that IPv6 works too, odd problems can sometimes be down to broken IPv6
The image is the install, why would you need an installer?
Then it has to download the containers, and those then have to download their dependencies
There's no way to do this 100% offline
Not without shipping outdated installs
Because at the pace that HA is growing, that could lead to tons of issues. People being stuck at some old version, etc and not matching current docs and such.
HA is meant to be a private smart home controller, not nessecarily an offline smart home controller.
There's an update every couple of weeks, if not more
Yeah its just frustrating being stuck on the "no" version
What do you mean the "no" version?
As in its currently not possible for me to install
Ah, but that's really outside of HA's control. I mean, if github sucks (and then often do in this area), there's really nothing that can be done about that. TBF, every smart home platform runs into this same thing. When you plug in a Smartthings hub for instance, first thing it does is reach out to the net to get stuff. If smartthings is having problems with their platform (which they often do), it's the same problem.
I think we made a lap here, that's why I was wishing for the ability to have an 'offline' install, because currently it can become uninstallable outside your control
but I understand
If unraid supports Docker you could try a pure Docker install, and see if that works
I had a docker install already, just deleted it to move to VM version because passing USB devices was being a pain, and is much easier with VMs on unraid. Also figured might as well do it to have the addons as well
All of this just so HA can know the humidity in my room 😅
Got a pi5 today, used pi imager like described in https://www.home-assistant.io/installation/raspberrypi
pi does not boot HAOS, then i used etcher and used the url for PI5 from here https://www.home-assistant.io/installation/raspberrypi#writing-the-image-with-balena-etcher.
Then i tried PI OS this is booting, did bootloader update and Boot order to SDCard.
Tried 2 different SD Cards.
Tried both ways PI Imager and ETcher with both SDCards.
Pi says everytime
Boot Mode: USB-MSD (04) order F4
USB2 ...connected
USB2 ...connected enabled
USB2 root HUB port 2 init
USB MSD stopped. Timeout: 25 seconds
Boot Mode: restart (0f) order 0
Boot Mode: SD (01) order f4
trying Partition: 0
trying Partition: 0
trying Partition: 0
trying Partition: 0
and then he retries in a loop
What does it display?
but raspberry pi imager is not working too. and rasp OS is running.
@brave thorn I converted your message into a file since it's above 15 lines :+1:
completed my post
FWIW I just spun up a HAOS VM in VBox and it installed fine, so if you're still having issues it's something between you and GitHub, or with your VM/network. Maybe try a VBox test install?
i have a running haos on a x86 device, i want to move to the pi5 (power consumation and the device can do so much more) and i need the addons for my current enviroment
sorry new on discord
@brave thorn 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).
downloaded a old verions 12.2, wrote it to sd card with balina etcher and the pi5 was booting perfectly
If Home Assistant boots from an M.2 NVME drive on say an RPi5, does it automatically detect that and enable a longer history retention or does that still have to be set up manually?
You still have to setup the recorder integration for longer retention.
I too am finding problems putting this on a BeeLink NUC. I'm suspecting the image! I have downloaded it twice. I get a GNU GRUB Version 2.12 screen, and then 2 primary and 2 secondary choices of boot image. Nothing happens on any of them. When I flashed the image to the 256GB SSD, BalenaEtcher flagged an error at the end of the verify image phase. I downloaded a second copy of the image, and tried flashing it to a Samsun 1TB drive that I have used before. B.E. then gave this error "Something went wrong while opening "C:\Downloads\haos_generic-x85-64-12.4(1)img.zx"" I also tried to go the Ubantu route, but ran into the problem of boot choice. The BeeLink won't allow me to put the Sandisk drive into the top boot order position. (The drive is good as I tried it in a
Try extracting the .xz yourself and give it that instead or try the pi flasher.
You could also virtualize HAOS which can make a lot of sense on such a PC.
x85
😄
I too am finding it difficult to install the HAOS image on my BeeLink NUC. I tried the Ubantu method, only to find that I couldn't put the Sandisk stick high enough in the boot order to be found. I tried the same stick on a GM1Ktec NUC and had no trouble booting into Ubantu (at a very high resolution of more than 3K x 2K that I wasn't aware it supported) Back to the image being fashed to a 256Gb SSD. It always give an error at the end of the verify phase in BalenaEtcher. I downloaded a second image and it gave me this error: "Something went wrong while opening C:\Downloads\haos_generic-x86-64=1.4(1).img.zx" I tried the operation from the URL, and it said the image was corrupt. Someone please tell the image is good, that I am a fool, and all I have to do is???????????????
I told you above. Etcher seems to have a moment. Older versions supposedly work.
Can you provide a picture of the boot menu in the UEFI of the BeeLink?
So is it better to boot from the MicroSD and then set up the recorder integration for longer retention storing onto the M.2?
Or should it just be booted directly from the M.2 from the start?
The problem is I don't think I have an easy way of flashing an image to an M.2 drive with any sort of ease, so maybe easier to do the former.
M.2 cases or docks can be had for 10-20.
Yeah but usually they're for SATA M.2 which works fine on USB
PCIe M.2 really needs Thunderbolt
Which I have in my laptop but I don't really want another M.2 dock 🙃
Look for one that can do both.
question, installed a new HA on proxmox. But i cant update the supervisor version to the current vers.
Command completed successfully.
➜ ~ supervisor update
zsh: command not found: supervisor
➜ ~ ha supervisor update
Processing... Done.
Error: No supervisor update available - 2024.06.0
Supervisor 2024.06.0 is the present stable version
2024.6.4
2024.6.4
12.4
12.4
2024.06.0
2024.06.2
it wont let me update to the latest supervisor?
i cant restore my backup:
Backup was made on supervisor version 2024.06.2, can't restore on 2024.06.0. Must update supervisor first.
Error: No supervisor update available - 2024.06.0
➜ ~ ha supervisor update --version 2024.06.2
Processing... Done.
Error: No supervisor update available - 2024.06.0
➜ ~
Then you have to switch to the beta branch
Settings - System - Updates -
menu - Join beta (or sth. like that)
That was it. Thank you
So I had to do a server reset on HA, and when it came back, my zigbee topology is a bit broken. I have one device that isn't linking into the mesh, and took one other device with it. How would I re-pair that device to the mesh without breaking all the tracking that I have on it?
https://imgur.com/a/Kilsada as close a view of the topology as I could get
I let it retry all day, never worked. No way its githubs end or you'd have other people having this issue by now, no? Any idea how to troubleshoot it?
Hey all! I was told to post here. I'm running the official VM image on a debian bookworm host with qemu-kvm / libvert. It has been running fine for a year with the exception of it clinging to a very old DHCP lease (for months!) and not renewing it. It finally dropped the lease today, and now the interface down; in auto config, but no lease - all the octet fields are blank. I'm baffled because everything else had been working fine (accessing it from a .local address, esphome which uses mdns), everything works on the host, including ~2 dozen docker containers.
Previously trying to troubleshoot this I'd toggled the interface between manual config and DHCP both in the web UI and via CLI using ha network.
Where can I find logs relevant to the dhcp client, and what other troubleshooting steps do people suggest? Google hasn't helped much
Additional note: ran nmcli to edit the Supervisor entry, did save and activate, printed something about waiting for DHCP, finished in barely a second or two...interface still wasn't configured
Set a manual IP address and still no network access - except I'm able to access the web UI. But updates fail, esphome isn't working, weather and smartplug stuff is also broken, etc.
Is there anyway I can just download whatever it's struggling with and transfer? Or mitm this?
Should I make a post on forums?
Yes
U are not bocking https ?
If ping is ok then it should be https...
Internal Server Error ("Get "https://ghcr.io/v2/": net/http: request canceled while waiting for connection (Client.Timeout exceeded while awaiting headers)")
no proxi, firewall, pi-hole... ?
manual IP also requires DNS entry
U can access the webui without network ? (hum) ...i missed something....
I was speaking in shorthand - I gave a router, netmask, and DNS server. I can access the web UI, but home assistant is unable to reach anything on the network. It was a day ago, though DHCP was the only thing that didn't work.
HA is able to ping your dns ?
( may be restart your DNS server ? )
other VM are fine ?
@olive ember It eventually worked, no clue why
@vast sapphire Ah !!! NICE! .....:-) what have you done? restarted everything ( wifi, router, ...) ??????
Ive tried restarting it a bunch, tried changing to preview release, tried a lot. Wasnt doing anything when it finally worked, just left it up and after hours of retrying it worked eventually
I dont get it at all
DNS works fine on my network. It's not host connectivity (as I said in my first message, the host and 20 containers on it work fine) and it's not my network. I don't see an iptables rules that would be causing this. And no, I can't ping my DNS server - which is also my router - from the HA container. But that's the whole problem - homeassistant can't make any outgoing connections.
I can see the interface is configured correctly and there's a (correct) default route...still nothing. I can get into the web ui, nothing else.
it's the HAOS vm?
this is just me, but I would download the backup file from it, turn off the VM and set it aside, and then install a fresh VM and restore the backup to it
vms and containers are pretty disposable, everything that makes it special should be included in the backups
Please share a network diagram and/or description about your network and the output of ha network info.
@dense gate I converted your message into a file since it's above 15 lines :+1:
Please use a code share site to share code or logs, for example:
- https://dpaste.org/ (select YAML for the language, and consider picking a longer expiry)
- http://pastie.org/ (select YAML for the language)
- https://paste.debian.net/ (you guessed it, select YAML as the language)
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.
Well, for some value of readable - not at all accessible on mobile
How are you running the VM and where?
so, repeating the core question:
Where can I find logs relevant to the dhcp client, and what other troubleshooting steps do people suggest?
any of that would only be available via the CLI, through SSH or with a monitor & keyboard plugged in
It's a virtual machine.
i'd start with ha network info
doesn't matter, you need to get to the CLI
that's step 1
step 2 is running commands
You use a static ip. The connectivity problem is thus not really related to DHCP, no? Do other VMs work fine?
The logs can be checked via journalctl: https://docs.redhat.com/en/documentation/red_hat_enterprise_linux/8/html/configuring_and_managing_networking/introduction-to-networkmanager-debugging_configuring-and-managing-networking#debugging-levels-and-domains_introduction-to-networkmanager-debugging
You have to run this at the system level, not via addon. Enter login first. I recommend you set up developer SSH for copy paste support and such.
I'm able to access the console of the VM just fine, because it's a VM...I can access the HA docker container because I have the web UI terminal add-on installed. And in both my first message (linked to a few comments above) and in a follow up I mention using the CLI.
that's great, but that still doesn't tell me anything new
Note that you cannot run commands like journalctl in the addon since it's a docker container.
but they can run ha network info to see if HA can get out
but you can run shell from the console, that's fine
exactly.
I only just got told to look in journalctl and have been answering you all. I have been using the console to run ha, nmcli, etc...without success.
you're saying ha ... commands don't work?
and so you're wasting time again by replying instead of doing the thing.
we'll wait
I'm curious how the VM's networking config looks like.
I'm curious about this
I have been using the console to run ha, nmcli, etc...without success.
he means without success in resolving the problem
well they aren't to resolve the problem, they're to start identifying where the problem lies
I meant that I tried reconfiguring the IPv4 configuration and it did not fix the lack of connectivity
do you mean the output from nmcli in the VM?
Seven Minutes of following instructions
No the libvirt one.
ah, you mean the VM's config on the host system, not the VM's OS networking config. Got it. give me a bit.
Back, was checking some stuff, referencing this against HA KVM setup guides etc.
here's the VM's network interface config.
<interface type='bridge'> <mac address='52:54:00:ef:c6:ff'/> <source bridge='br0'/> <model type='virtio'/> <address type='pci' domain='0x0000' bus='0x01' slot='0x02' function='0x0'/> </interface>
Bridge is fine. I think it uses some weird NATing by default.
yeah, and this was all working fine for a year-ish, with the exception of DHCP never working beyond the first lease pull.
Do other VMs work fine? What does ip a ln the host say?
journalctl shows:
homeassistant NetworkManager[322]: <info> dhcp4 (enp1s2): activation: beginning transaction (timeout in 45 seconds)
homeassistant NetworkManager[322]: <info> dhcp4 (enp1s2): canceled DHCP transaction
homeassistant NetworkManager[322]: <info> dhcp4 (enp1s2): activation: beginning transaction (timeout in 45 seconds)
homeassistant NetworkManager[322]: <info> dhcp4 (enp1s2): state changed no lease
(removed timestamps for readability)
It's the only VM. 20 docker containers work fine. host has one bridge; physical interface is in the bridge, bridge has a static IP on the local network.
oh, correction: docker has its own bridge, sorry
I always use dhclient -v or tcpdump to debug DHCP but that doesn't work on HAOS. You can technically install and run tcpdump in the SSH addon but you need internet for that.
I'd either increase verbosity of NetworkManager if possible or clone the VM and use another OS to test. I'm not convinced it's a HAOS issue at this point.
Or even just make a snapshot and install debian over the current VM without modifications.
I'd still like to see ip a of the host.
HA URL not accessible
You can leave the Beta branch now. Supervisor 2024.06.2 has been released as stable now.
i am getting a access denied to the vdi
like
in the vm
everything else is fine
So in February 2021 a Home Assistant forum moderator said that recorder was a bad choice for recording long term data because it corrupts easily and eventually takes up a huge amount of space.
They then recommended influxdb instead.
Is this no longer the case?
There have been a LOT of improvements to the database, yeah. But if you are looking at storing years of data, InfluxDB or Prometheus are better suited.
I mean I don't think I need more than a year of data, but the more stable & less file size the database is (and ease of recall obviously) then the more chance of longer data retention.
True. Better on your backups as well. But, even for storing a year's worth of data, I'd still opt for a timeseries database. It's literally what they are built for. Plus, better slicing and dicing of your data as well.
In my experience it slows down a bit if the database goes above 3G or so.
Simply running the queries from here is quite agonizing: https://community.home-assistant.io/t/how-to-keep-your-recorder-database-size-under-control/295795
I use a whitelist based approach for that reason.
Oh, I forgot disabling pinging. My bad.
Yup. Same. My DB size never goes above a 1GB and if it does, I start to panic. lol
lmao you're good... I'm not Tinkerer 😉
So if you're using say Prometheus, how does that tie in with the "Show More" buttons showing you the regular sort of history from Home Assistant's recorder?
Presumably you have to just use Prometheus to browse that data?
Prometheus' database is not good for storing long term data either. I assume you meant Influx? The recorder does not tie in with it. It runs in parallel. I believe the docs mention that as well.
Prometheus is ok for long term storage, but not nearly as efficient as InfluxDB is.
It doesn't (as @raven forge said). But, you CAN create sensors from your InfluxDB data and use them in HA Or you can use Grafana.
So basically, unless you use the native recorder you have to poll another database to get historical data
Well, you use both. Recorder for recent history and then InfluxDB for longer or more specialized readings.
I can't seem to find it now but I believe their docs said something to that effect some time ago.
Prometheus has been trying to play catch up for years. About the only real benefit that I use is their instant vectors which you can't really do with Flux in InfluxDB.
Yes that's if you actually set up and use InfluxDB. If you just booted onto an M.2 drive to begin with and configured recorder to have a 365 day history then presumably you would just do the normal "Show More" procedure to find data for a sensor etc.
Note that the long term statistics are a thing so even if you use a retention of 10 days you might still have history: #general-archived message
In my opinion the HA graphs and tools are not really good to inspect long term data though. Especially not granular data.
With Influx and grafana you can do proper calculations and dynamic dashboards so your data can actually be used for something useful.
Sure. That's possible. But, you run the risk of corruption and a huge DB and performance impacts to HA. It really depends on what you want to actually do with the data. The question I typically ask users if why they need to know what the temperature was on a sensor back in 2022.
I have a dashboard for my contact sensors. I can choose a time range and it tells me the times it was triggered in total, average per minute/hour and so on. Is it useful? Probably not. But that's the kind of flexibility you have.
Just run Home Assistant on a 40 core 80 thread dual CPU server with 192GB of RAM and TB's of Flash Storage and Raid pools. Performance issue fixed. 🙃
I use VictoriaMetrics as time series database. It accepts data in Influx protocol (and others) format but uses MetricsQL to query data. Basically a fork of prometheus' PromQL. I can't deal with Influx's language. I also use it to scape. It replaced prometheus for me.
I've heard really good things about VM, but I have yet to really play with it (we are full prometheus and influxdb at work, so yeah). I'll have to setup an instance here and see if I like it.
Hi all,
I have a super strange problem: I have HA installed via HA-OS on a custom server. I need two WiFi interfaces (TLDR point to point connection to two solar inverters) . I have two identical USB-WiFI sticks.
If I connect either one it works.
If I connect both I see both in the hardware section but I only see one in the network section.
Any idea why?
Please share the result of this: #zigbee-archived message
I’m not sure how, you mean via the terminal add on?
If I check in the settings—>Hardware I see
WiFi 1 as /dev/bus/usb/001/006
WiFi 2 as /dev/bus/usb/001/012
And the strangest part: If I swap one of the identical USB-WiFi sticks for another different model that I borrowed of another device everything just works
It seems as if HA misbehaves with two USB WiFi Sticks of identical model
Yes via the CLI.
That is interesting: for one stick I get Driver=one and for the anther Driver=rtl8xxxu
I updated Supervisor from 06.0 to 06.2 and now HA can't start. I have access to the system but ha is down and I'm getting homeassistant hassio_supervisor[590]: s6-supervise supervisor: warning: unable to spawn ./run (waiting 60 seconds): Exec format error and homeassistant hassio_supervisor[590]: s6-supervise watchdog: warning: unable to spawn ./run (waiting 60 seconds): Exec format error. Can't find any useful info on what could be the cause.
Exec format error usually means you've pulled a container for the wrong CPU arch
That told me exactly nothing - sorry 😄 It's not docker installation if that's what you're suggesting.
Well, you said you pudated the Supervisor, so you've got Docker 😉
How did you update?
Through the UI. There was monit that there's an update: Home Assistant Supervisor Update
Are you gonna share the output? Also try running ha host logs -vf and then re-plug the stick without a driver.
It even had link to the release notes: https://github.com/home-assistant/supervisor/releases/tag/2024.06.2
actually Im trying to copy the output to paste it somewhere but I somehow cannot ☺️
Please use a code share site to share code or logs, for example:
- https://dpaste.org/ (select YAML for the language, and consider picking a longer expiry)
- http://pastie.org/ (select YAML for the language)
- https://paste.debian.net/ (you guessed it, select YAML as the language)
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.
Use a proper SSH client or see here: #installation-archived message
Hopefully you use HAOS, in which case
ha supervisor repair
``` over SSH will probably help
Seems like their supervisor is broken through.
Well...ha doesn't work.
Running ha gives: Error response from daemon: container 4965af3ea267fc62f6884b702fba1033ed1430c0a78397efb3ff6921eac397da is not running
Reboot and pray?
The fastest way if that doesn't work is to re-install and restore. I walked someone through re-installing just supervisor here so it's possible: #1228804362231349409 message
You need to do this via keyboard and monitor through and it might not work in your case.
Already did rebooting...and praying - didn't help. 😉
from HA-->Settings-->Hardware I can see that the two USB Wifi Adapters are on Port 4 Device 006 and 012
Yup I also see two devices using rtl8xxxu at those ports.
yep currenly for testing I have three WiFi Adapters on it,
Two identical ones 006 and 012 and another one of another brand 009.
I see all in the HA Settings but ontly two show up as NIC in the Settings-->Network
So what did ha host logs -vf say when you re-plugged 009?
Well...your trick worked! Thank you! It took quite a while to restart - command worked relatively fast but then all HA processes restarted, which for the moment looked like the trick did nothing.
I should mention that the aarch64 has to be changed to what your device uses.
docker ps -a should show the arches for the other (or existing supervisor) container(s). Or just uname -m.
Well...I should mention what HW I'm using, which is Pi4...And I hope I didn't broke anything more than it already was. 😄
I didn't do any reboot on my own. It simply restarted processes from what I've manage to get from journal. I can try to do the full system reboot, if that's what your advise is.
If everything works fine right now I'd run a ha supervisor repair, reboot and see if everything is fine.
If you don't have backups take some. I guess you can also just leave it as is.
so I did the following:
- Attached two DIFFERENT WiFi Sticks
- Fresh boot of HA --> everything works as expected
- ha host logs -vf
- Attached the third stick which is identical Hardware as one of the first two
WOW NOW it works!
strange, seems something is colliding with the HA boots up with two identical sticks....
Got: Post "http://supervisor/supervisor/repair": context deadline exceeded (Client.Timeout exceeded while awaiting headers) after running repair. I'll try to reboot and see what will happen. It's better to get issues now, when I can do anything about it then when I'm not around. 😉
Got this message while installing HAOS on odroid N2+. The installation does not go beyond setup page. Any pointer to what could be the problem.
Likely network/DNS issues.
Where does one find assistance with Nabucasa ... trying ot figure out how to use a custom domain name with cloudflare. I can validate the host - but the acme_challenge won't validate...
There’s NC support, otherwise #cloud-archived
i just installed my ha on a vm and imported my profile that was running in a docker container it also had hacs and so on anything i need to be aware of or that i shuld reinstall or check ?
So seem to be stuck trying to install 12.3 on a rasp pi 5 because 12.4 is broken. Also seems to get stuck trying to update the supervisor so it wont finish startup....any clues where to start or look for help?
I'm helping a friend setup a sff i5-6500 system to run virtual machines and home assistant.
What is the recommend os to run HA OS and docker?
You aware aware that HAOS is a OS and does not support running random docker containers? For virtualizing I recommend Proxmox VE which is debian based.
I recommend debian as OS for your VM running docker as well.
That's why I mention ha os and docker.
So proxmox VE Is what most people use?
I don't know what most people use but PVE is pretty popular from what I can tell and it's what I use myself.
For some odd reason after last ha core update my HA is really slow & sometimes unresponsive.. I restored a backup but after updating HA Core its really slow & unresponsive again
In running haos on a Nuc
You might want to do some more debugging. You've provided us with 0 information aside from hardware. Are there errors in your logs? What's slow and unresponsive? Automations? UI? APIs? Addons?
Hi guys,
I plan to install home assistant on my synology nas. I read on official website that container was preferred in that case but I have a Z-wave stick and read that everytime the nas reboots, the sticks becomes unaccessible and we have to replug it
I also have virtual machine manager on my nas, is it not better to use the ova image given ?
As long as it supports USB passthrough then VM makes more sense in that case
IIRC, there are issues with USB passthrough with the Synology and VMs.
IIRC, there are issues with
USB passthrough with theSynology
FTFY
lmao that too
what do you mean by USB Passthrough ? Currently I have a jeedom vm installed and everything is ok
even if the nas reboots it gets back access to usb stick
IIRC, it affected certain USB devices (mostly zigbee/zwave sticks) that were not getting passed through to /dev/ in DSM.
mmm ok .. in my dsm it's ok
as long as it's configured in the vm
and it's resilient
when my nas reboots jeedom reboots and still has access to the stick
Ok, that's good. Ideally, running HA on a Synology should be considered a temporary solution, imho.
I just wanna test for now
and see if I don't lose anything and switch
I mainly use z-wave and alexa vocal assistant
I saw there was a skill on alexa for HA
There is a skill for HA, but it leverages Nabu Casa.
But yeah, trying it out is good. 🙂
i saw there was another solution without nabu casa
There is. You can manually create the skill in AWS and do it that way as well. Nabu Casa is the dead simple approach, the manual way is a bit harder.
i'm a hard player lol
lmao well, there ya go!
Yup. That's a good integration. Pairs nicely with the Alexa Media Player integration as well.
UI & automations are slow i can show my logs
@neat forge I converted your message into a file since it's above 15 lines :+1:
i think this might be something
Please report it to the author of the 'roborock' custom integration.
So ive been running an overly powerful mini pc with Windows for a few years now and its only really used for HAOS running in a VM on it. It's nice because im highly fluent with the Windows OS and performing full Image backups, individual VM backups, snapshots before monthly updates etc. My plan has always been if any HA related HA upgrade goes wrong, a simple snapshot restoration immediately resolves. If a full host hardware situation happens i can offload the full VM backup to a temporary second machine with VMWare workstation and literally boot up my HAOS VM instance again. And even in a full host hardware failire i can universally restore the entire hosts Windows OS image to totally different hardware to be fully up and running.
At some point i want to proceed with a proxmox high availability concept with a 3 node HA system but i have work to do there and stuff to replan out. I was playing with the idea of taking Windows out of the equation and installing the HAOS directly on the mini pc hardware. Is it safe to say if i perform a fresh installation of HAOS on said mini pc directly, i can simply take a full HAOS backup that i would of took before hand and restore that directly and everything would mesh like it exactly was when HAOS was running in my VM?
Yes.
(Sorry, that was a bit succint, but yeah, you can restore a full backup to a new instance and it'll restore everything).
great, thanks!
Hello, i've a small question :
I just lost all my backup from a old HA installation (HAOS 5 i think).
I will begin a fresh new installation, is it easy to link again all my sensor already placed in my house ? It's zigbee aqara sensor.
I use a raspberry 3b+ and conbee usb
You mean because the device associations are present on your coordinator, but you no longer have the entities in HA directly?
for exemple, the door sensor is fixed to the door already ^^
I mean, easy-ish? Depends on how many sensors you have and such.
7 sensors fixed on doors and windows
for the other temp sensor they are not fixed so it's easy
in my memory it was needed to put the sensor close to the conbee to pair it
What do people do in regards to coordinator backups? Like i believe either the Zigbee or ZWave addon i know you can do "radio" backups, but one of them i didnt see a backup function, just forget which. And im assuming the file backup\restore concept is exactly that easy. So if your coordinator burns up, dies, whatever and you get a new one, you clone it over and the devices know no different. But if thats not the case, then what is the process of basically getting all your devices synced again in this disasterous scenario? Is that all a brand new\start fresh effort with all your devices?
I have like 57 zigbee devices and mine have all paired fine with my conbee in my basement in a 3000 sq\ft house.
So maybe depends on your signal saturation?
i will test now, just finished the fresh install
using ZHA ? In past i was using deconz, but idk if it's still good
At least you will be up-to-date 😉 ... and do some backups ?
deCONZ is very much the poorest #zigbee-archived option, ZHA or Zigbee2MQTT (which you use is largely personal preference) are better choices
I tried setting up https using lets encrpyt and it seems to work, but after about a month now I am having the issue that on my PC I can only acsess it using its IP (not homeassistant.local) and I cannot seem to acess it from other devices on my network
I am also getting this error in the network tab Invalid local network URL You have configured an HTTPS certificate in Home Assistant. This means that your internal URL needs to be set to a domain covered by the certficate. I have tried restarting and had a quick google to try and figure out what is wrong but I can't seem to get it
I have a suspicion that it is to do with some mistake that I made setting up the HTTPS but I honsetly have no clue
So... you have to use https:// all the time now
If you don't use the hostname in the cert then you'll get errors, and many things will refuse to connect
Have you:
- Done what the error above told you to do?
- Tried connecting only using the hostname in the cert?
Okay, I think I am beginning to understand. I initally did all this so that I could use homeassistant with google assistant but I don't know if it is worth the hassle
- I don't think I completely understand what it is saying. If I update the interal URLs to be within the domain of the domain in the cert does that mean that I would no longer have a true interal url as all traffic would be going through the external domain?
- If I try to connect to homeassistant.mydomain.com or homeassistant.mydomain.com:8123 it says it's unable to connect. It's been so long since I did it this way, is that the correct way to do it?
- Sort of, but it's just going to your router and back
- Because you have to use
https://
https://homeassistant.example.com
Assuming you forwarded 443 on the router to 8123 on HA
If you want internal access without https:// then you need to use a reverse proxy (eg the NPM add-on, or Traefik if you use Docker) to handle SSL and remote access
Wait I think that port forwarding might have been the issue for why I could not acess it with HTTPS. I have now done that and ```https://homeassistant.example.com
This isn't really an issue now that that is sorted but it still puzzles me why I am able to access it using it's IP directly but only from my PC and from no other devices
Thanks again for your help
I try use Nginx Proxy Manager
When add host then error Internal Error
As always, check HA's log, check the proxy's log
I'm having issues with getting HA running on Proxmox, are those questions allowed here?
I had my install working with https/letsencrypt where it was getting a public address and was able to http-01 validate. I lost that address and need to proxy to this instance locally... can I easily restore just using http with the instance? This is a rasberrypi install, I can ping and presumably ssh into the instance... but it doesn't seem to be allowing me to do http to it via the browser.
A lot of folks run on Proxmox, what's the issue?
I used the script by tteck to create the HA instance, but it doesn't seem to be coming up properly. Here's the output from the console when I try a few various commands: https://i.imgur.com/Zd7SKH8.png
No response from the web UI. Any other startup logs I can check?
Hey guys i wanted to ask if there is a good tutorial where it is explained how to install HS on debian 12 lite and use it in kiosk mode? i have done some tuts now but none i found works. On pi3 its for development for a psb board
Hello everyone, an update went wrong on my HA, and I can no longer boot my server. My backup was old so I can't even say to recreate an image.. Could you help me? I'm blocked in sheel.. I had spent weeks configuring my server.. Thx you in advance
You'll have to be more specific about what it's doing, what messages you get, etc. Is it connected to a monitor? Can you see the CLI? What hardware? What type of installation?
I'm on a VM with a qcow2 disk type, and I get the UEFI window with the Sheel when I get to HA. I have the impression that the partition is corrupted following an update that went wrong, however I would like to be able to recover my data or do something.. thank you!
Hey guys i wanted to ask if there is a
A couple of days ago I posted about problems with the haos qemu container that had been running OK for ~1y, with the exception of the DHCP client not working. In the process of setting up a new VM to restore a backup into, I came across the community instructions again, and there was this little note under the "disable firewalling on the bridge" section:
NOTE There is another alternative suggested by @brady in this post where you instead tell Docker that the bridge exists.
...and that fixed my networking problem. I think the setup had been partially working because I had some iptables overrides in a sysctl file, but for some odd reason they didn't match the ones I see suggested in the above thread. I didn't realize what was going on becaus apparently I can't read an iptables output - it looked OK. Time to refresh that memory, I guess.
Why this broke days after a docker update, I'm not certain. System wasn't rebooted or anything else obvious.
So for future reference, if someone says their QEMU guest has no network access and they mention they're using docker, it's likely something happened to whatever was set up around stopping docker from firewalling their bridge.
I have HA running in proxmox on a computer with a built in bluetooth adapter. how do I pass through that adapter so that HA recognizes it?
oh. since it's probably connected via usb, I add a usb device?
yup. k, thx
Ok, this is going to be a bit confusing to explain for me given im new to everything networking and home assistant so bear with me here. I have two separate locations i want to monitor with home assistant, one is my home and the other is a hangar. there arnt many devices at the hangar, mainly light switches, motion sensors, cameras, and ac/heating. i use ubiquitis unifi system and both locations have ubiquiti routers and they are connected using a thing called site magic. which apparently makes them like share a network, paking devices accessable from one location to another. should i enable this feature and not add another home assistant instance for the other network or should i just let the devices be visible to my current home assistant instance. i already am adding a server at the hangar anyways because i want multile location backups for my nas. thoughts?
Hi @toxic maple . if your network is rock solid, then use only one instance of HomeAssistant... what about zigbee network? extended by ubiquity site magic ?
I can use a second signed coordinator. I found one that is Poe
Thank your for the help @olive ember
So on a direct baremetal installation of HAOS, how does one recover if a HAOS updates goes wrong? With a Core update, you can use the backup\restore. But if you did a full HAOS update and the system cant fully boot back up. Whats the recovery\rollback look like?
Reinstall HAOS and then restore your full backup (that you have stored on another machine).
Re-installing HAOS and recovering from a full backup (which you have hopefully downloaded or stored on another machine)

Makes sense. Is there a faster process for that other then booting a live linux flash drive..restoring a HAOS disk image...going through setup\restore again...
Outside of doing all this in a virtualized environment of couse...
VM is best, obvs. But no, there’s no faster process.
Ok. I basically been running an overpowered full pc with Windows, running VMWare and my HAOS VM for aobut 2 years. THat PCs starting to give me issues. I loved the flexibility of SNapshots\vm backups everytime i would update anything to do with HA. Been jumpin around with what to try next. I ended up getting one of those N100 minipc's to use as a dedicated HA box. Was going to go bare metal HAOS install on it, letting it be its own dedicated HA system. But then i was thinking how id be losing out on my whole easy backup\snapshot strategy and stuff like i had with VMware Workstation. So now im thinking Proxmox on the minipc and HAOS as a VM on that to gain that snapshot\backup\restore resilency back...but then im back to basically doing what i was before lol
lol you are, but honestly, for resiliency and ease, it’s the right choice.
I dunno man, I think the backup that HA provides is really good stuff. WHy would you need anymore?
Reinstall HAOS and click on the backup....so easy....
self inflicted frustations!! lol
VM gives you better backup options plus snapshotting and such.
OR save time increments....it's really well done.
YEah but you can do all that with HA backups and automations if you wish.
True, but it’s usually slower and you still need a way to export the backups. If you don’t want to use SAMBA (because SAMBA is crap), VM is the way to go. Plus, if you have good hardware and want to do other things with it.
I would LOVE to install HAOS bare metal right on the minipc and let that be all it does. It runs 24\7 never gives any problems. Doing periodic HA Core updates and creating backups and being able to restore them as a failsafe isnt my concern. Its when a HAOS update happens like this latest 12.4, and then your minipc fails to boot in a catastrophic way...now you cant even get to the HA instance to do anything. Youd find yourself with much longer downtime needing to restore a whole HAOS disk image from a linix live boot scenario, etc etc.
I mean, I don’t run HAOS, so I don’t use backups at all, if I did, it would be using a VM.
You dont need samba, even an attached drive would do. I think they nailed it with the back ups.
Ha updates make a backup snapshot automatically now
Yeah, but I don’t want attached drives that can fail. I have a NAS server with RAID plus offsite backup. Yes, all of that can be done in HAOS, but that’s a single point of failure that I wouldn’t want.
Plus, with a VM, I can spin up multiple instances easily to do things like test #beta builds. 😉
I must be missing something. If you click Install on a HAOS Update, like 12.4 right now. And the whole system goes down and never comes back up, the CLI cant load, etc, how are you revertin to anything
I would keep all betas on a separate machine, and not touch you main production with anything beta.
And if you don’t have extra machines? What then?
You gotta make due with what you have I suppose.
You dont have extra machines fore your extra machines!?? lol
@timber lagoon I’m not saying you’re wrong (you’re not), just preferences.
lol on the contrary, I have too many machines. Hahaha
YOu reinstall HAOS and at install just point to the backup file home assistant made before the backup
Thats my point, and in order to do that first part....you need to boot a linux live usb to access Disk utility, to point to a haos_generic-x86-64-12.4.img file to restore to the disk, to start over..... compared to 45 seconds of creating a virtual snapshot and being able to restore it if it all went south. Not arguing, having healthy debate as i still juggle with what way i want to go here.
NO you dont need a linux live boot usb...lol
Then what boot medium are you using to perform disk image restorations?
HA install will use it during set up
That's why I recommend virtualizing HAOS to anyone willing to listen.
you can literally plug in the HA usb image, say use this backup image and it boots up like nothing happened
That’s the point though, you still need to keep that boot medium somewhere. With a VM, you can spin up a new instance from a snapshot in a few seconds and then restore to that.
Fair enough. Anyhow, moving on. Both methods work. How you choose to approach it is up to you @dreamy berry
YEa, thats all fair. I think i might still be missing something that youre saying @timber lagoon and i want to understand. Im gonna type up a big scenario to see where im probabably inaccurately describing something. But it sounds like youre saying something is a bit more resilient there that im unaware of. Please gimmie some time while i type something up easily digestable and if you wouldnt mind, can hlep me fill the missing gap.
In the mean time, my next topic for the day. PRobably better in ZigBee or Hardware perhaps. But maybe you fine gents would know regardless. I know theres interference problems with USB3.0 ports\devices when it comes to ZigBee communication. I know the easy answer is use USB2.0 ports for your coordinator. Along with use a USB extension cable if you need to use a USB3.0 port. Does USB3.1 or 3.2 ports share the same problem? Or did they change the encoding\RF stuff there after it was acknowledged with USB3.0?
Same issue. USB3.x is a leading cause of cancer… erm, interference, for Zigbee. 😊
Anyone install Home Assistant, Mosquitto, and NodeRed on Docker-Compose, but Mosquitto container gets stuck in the starting phase?
I'm stumped
Share your compose file and log. Doesn't sound related to HA though.
Please use a code share site to share code or logs, for example:
- https://dpaste.org/ (select YAML for the language, and consider picking a longer expiry)
- http://pastie.org/ (select YAML for the language)
- https://paste.debian.net/ (you guessed it, select YAML as the language)
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.
Is one of these methods better then the other?
USB3.2 Port > 6ft usb extension > zigbee dongle
USB3.2 Port > 6ft usb extension > usb2.0 hub > zigbee dongle
USB3.2 Port > usb 2.0hub > 6ft usb extension > zigbee dongle
I'm not sure if going through a USB 2 HUB helps anything. It's not like it would filter anything, right?
I followed this guide, one sec and I'll post my YAML
Not based on these comments.
Issue: USB 3.0 ports/computers/peripherals are known culprits of RFI/EMI/EMF disruption. (See Ref. 1 23 and 2 9)
Action: Make sure to only connect the Zigbee USB adapter to a USB 2.0 port (and not to a USB 3.x port)
Tips:
If your computer only has a USB 3.x port then buy and connect the adapter via a powered USB 2.0 hub
Using via USB 2.0 hub will in practice convert USB 3.0 to a USB 2.0 port and thus avoid USB 3.0 EMF
A USB 2.0 hub that uses an external power supply will ensure power requirements are fulfilled
Source?
Also see the link here: #the-water-cooler message
That's my docker-compose.yml
mosquitto gets stuck starting, waited 1200sec, and I have nodered as a dependency on mosquitto, so that isn't starting without mosquitto
Please share the full mosquitto docker log.
REading bunch of different articles, but got that specifically from this one. https://community.home-assistant.io/t/zigbee-networks-how-to-guide-for-avoiding-interference-optimize-using-zigbee-router-devices-repeaters-extenders-to-get-a-stable-mesh-network-with-best-possible-range-and-coverage/515752
I need a better source to believe with certainty that passing through a USB 2 hub helps anything.
No logs created, but there wasn't a log directory, so made that and trying again to see if the logs populate
Try docker copose down. Then run docker compose up mosquitto and see what happens.
So your understanding then is the extension cable in the means of its just getting your coordinator AWAY from any USB3.x port and\or other device plugged into one is really wants preventing it from being affected from the given off interference. Aka distance is solving the problem, nothing more.
Error response from daemon: driver failed programming external connectivity on endpoint mosquitto (69e49818fee1ed64479f23e32c77c61d97dac736dacd10b30f554d4c525d50b5): Error starting userland proxy: listen tcp4 0.0.0.0:1883: bind: address already in use```
I'm using 1883 for something else?
I'll try swapping
I recently switched to a SLZB-06 but before that I used this janky setup with a Sonoff P on a USB 3 port which worked fine: #zigbee-archived message
Intel says the interference goes through the wire but how much and how far it travels I don't know. Does it travel through the data pins? The power pins? All of them? I can't imagine a USB 2 switch would filter this. At the end of the day I do not have enough information to tell anything for certain. I did not do any tests myself. I can only tell you that I had no problem with my setup.
I checked 1883, and nothing is assigned to it (at least that I can connect to), I changed to 1884, and when I compose up, it gets stuck on attaching to mosquitto
Can you try without the user:?
THanks. Im planning on going PoE based too next. Too many things i want to convert\redo at once. Trying to go slowly and non-frustratingly. No idea what a total Zigbee migration even looks like between coordinators. Hoping there no redoing of devices in anway.
Think I got it!
I think the dependencies were stopping it from running alone
So I ran the full compose up and it appears to be working (from initial looks)
Appreciate the help, @raven forge
mosquitto itself has no dependencies though so docker compose up mosquitto should work 🤷
We'll see if I run into any issues while I do the setup
I'm in the HA UI and doing the nodered stuff rn
Hi everyone. I hope this is the right place to ask for help. For some reason a few months ago my HA stopped updating and I now can't update the core or OS because something seems to have gone wrong with the supervisor update. I get the following error. Could anyone shed some light on this or maybe point me to how I might diagnose this?
Error: Update of Supervisor failed: Can't install ghcr.io/home-assistant/amd64-hassio-supervisor:2024.06.2: 404 Client Error for http+docker://localhost/vl.44/images/ghcr.io/home-assistant/amd64-hassio-supervisor:2024.06.2/json: Not Found ("No such image:ghcr.io/home-assistant/amd64-hassio-supervisor:2024.06.2")
im trying to setup mqtt on a raspberry pi 5 and i am getting told that i can only run it on a raspberry pi (despite me being on one) anyone know how to fix?
im trying to setup mqtt on a raspberry
is home assistant still working on the Raspberry pi 3B?
Yes
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.
Net: No Ethernet Found
but when i put in the other micro SD card its boots fine and picks up the ethernet
Did you choose the correct RPi model when flashing the SD card?
yeah
The one for RPi3, not RPi4 or 5
yes
Strange. Any DNS stuff like an ad-blocker running on your network? Or other restrictions in your router?
404 not found
Nope
ok theres lik 6 links what one doe i need to use lol its been forever since ive used imgur
not that i know of with my router
whats wierd is when i put the micro sd card in that has the Raspberry pi os on it, it picks up the ethernet
yeah
i have a 32 GB the 64GB that i bought specifically for the HAOS and the 16 GB one that had the RPi3 OS on it and they all say that same thing I've searched the NET: NO ETHERNET FOUND and it seems that others have had the same issue but i haven't found a work around or a fix that's why i came here to see if anyone knew of a way to fix it or knew what the problem was
Yeah. Maybe one of the RPi 3 users knows
Can someone help? I am installing HAOS on my old lenovo g780 and i am getting an error. There is a lot of red on the logs
Nvm resolved itself
If I install HAOS on my new Mini-pc does it run some sort of normal Linux I can also install other stuff on?
Or would I need to install something like Debian and go the docker route if I want to use it for multiple things?
Correct. HAOS takes over. You could also go the VM route as well with Proxmox (or other VM host) if you want to continue using HAOS but have other things as well.
I would recommend the VM way too. Even if you don't use HAOS. Here's a quick start guide for HA on Proxmox VE and some benefits of using VMs: #hardware-archived message
Hi there i got a mini pc i would like to sacrifice to the HA gods how can i do that
Ok so im not very tech savvy but from what i understand from that link i only need home assistant OS using Ubuntu
the container and core are other insulation methods right ?
Eh, you don’t even need Ubuntu. You can use whatever PC you have now to flash the image to a thumb drive or whatever media you have (even the drive on the device itself). When you boot up your device, it’ll automatically install everything you need. Super simple.
Yes, there’s Core (difficult), Container (easier) and Supervised (extremely difficult).
I didn't read anywhere over there that you could do a direct install using a usb maybe i skipped something
Ah, no. You’re right. I was wrong. Yeah, you can flash it directly to the drive you are using. The live USB method uses Ubuntu. Sorry about that!
Thanks for the help @proven grove i will work on it first thing in the morning
Sounds good! Come back if you run into issues or need more help.
I'm just sad i wasted my money on that linknlink gateway that was my stupidest purchase ever
I just got it and its buggy slow and very limited
I have a couple drawers full of stupid decisions. lol it happens to all of us.
While installing home assistant using the ubunto flash drive i reached home assistant CLI to be ready
CLI not starting jump into emergency console
Nevermind it just worked
Now i got to a page with some links my phone is searching for it for a while now but couldn't find it. Also i tried the provided links and didn't work
because of some unaxplained files in my /config directory (apparently remnants from times gone by, I also notice this strange difference on my .db file (SQLite default)
this is finder, and the seecond is in VScode
so, the VScode does not see any of the regular db files, but has a corrupt db, which I dont have in finder.... what the heck is happening here?
Some #add-ons-archived hide things to stop you fucking it up
If you search you'll see that discussed for VSCode
great, thanks!
and about the corrupt db? any reason I can not see that in regular finder (it isnt hidden after all with the .
Don't know how Finder works, but maybe the Samba add-on hides files too?
hmm, I can see all storage files etc, but yes, there were odd single files I couldnt see in the finder app either, so it's a bit of a guess , but Ill check Samba settings. anyways, should be safe to delete the corrupt db shouldnt it.....
It may be hidden in the code 
Hi.
I recently installed home assistant in docker on my ubuntu server.
I then started costumizing it but have reached the conclusion that every time my server goes down, home assistant is completely reset and i have to create a home and costumize it again.
Any idea why this is happening?
is it supposed to happen (do i have to save it manually) ?
Thanks in advance
Did you forget to mount the persistant HA config directory to the HA container?
docker run -d
--name homeassistant
--privileged
--restart=unless-stopped
-e TZ=MY_TIME_ZONE
-v /PATH_TO_YOUR_CONFIG:/config
-v /run/dbus:/run/dbus:ro
--network=host
ghcr.io/home-assistant/home-assistant:stable
The part-v /PATH_TO_YOUR_CONFIG:/config \, which has to be adjusted to your setup
A Docker Compose file would be similar
Both are not simply copy&paste from the website
yeahhh i dont really know about docker or docker compose
i will research more about this
that could be it but i think i did change it
docker inspect is your friend, well, compose files are your real friend
docker inspect -f '{{ .Mounts }}' homeassistant
I'd bet on you having messed up the mount/volume
hi there... use HA Container... it worked flawless for months, now I updated something
and HA doesn't boot anymore
HA starts in recovery mode with this error:
2024-07-01 17:53:19.137 ERROR (MainThread) [homeassistant.components.http.forwarded] Received X-Forwarded-For header from an untrusted proxy 172.20.0.5
I have this in my configuration.yaml
# Uncomment this if you are using SSL/TLS, running in Docker container, etc.
http:
server_port: 8123
use_x_forwarded_for: true # How to extract client IP address
trusted_proxies: # Permit connections from reverse proxy
- 172.16.0.0/12
- 192.168.0.0/16
it always worked well before...
I tried adding 172.0.0.0/24 but it doesn't work neither... 😦
tried 0.0.0.0/0 but it doesn't work...
I use HA since years and I destroyed it can't access anymore... any idea?
What the heck am I doing wrong? This worked for about 4 months until decided to redeploy HA, using the same version, 2024.3. HA behind Traefik.
use_x_forwarded_for: true
trusted_proxies:
- 10.0.0.0/8```
```←[31m2024-07-01 18:12:11.848 ERROR (MainThread) [homeassistant.components.http.forwarded] A request from a reverse proxy was received from 10.42.1.86, but your HTTP integration is not set-up for reverse proxies←[0m```
[supervisor.docker.interface] Can't install ghcr.io/home-assistant/qemuarm-64-homeassistant:2024.6.4: 404 Client Error for http+docker://localhost/v1.46/images/ghcr.io/home-assistant/qemuarm-64-homeassistant:2024.6.4/json: Not Found ("No such image: ghcr.io/home-assistant/qemuarm-64-homeassistant:2024.6.4")
i m inside a incus vm on a orange pi 5 plus
What the heck am I doing wrong? This
Make sure you actually made a directory (mkdir) that matches the pathway you specify in your compose.
Thanks for the responses @fallen imp , @tender wren
i reinstalled docker and compose and configured it again just to be sure and it all works perfectly now
In what sense?
Also check that you're not out of disk space
Text ... easy to read for everybody
Images ... maybe not easy
However, check why your system is unhealthy
how?
Should be somewhere in the system health page?
Open your Home Assistant instance and show information about your system
Maybe there? I don't use HAOS myself
i check
(and hopefully that's HAOS and not Supervised)
Click it, it'll tell you why
Somewhere it tells you why
Maybe there?
oh yes i clicked it
it says: The installation is not intact
Performing a broken installation will cause problems. Below is a list of problems you are having with your installation, click the links to find out how to fix the problems.
-supervisor
oh wait, i was cheching the official ha post of this post, and it says to update it manually
i cant update it manually because if i try to download the terminal ewxtension it says: Impossible to install the extension: AddonManager.install' blocked from execution, system is not healthy - supervisor
@humble mirage
@warped nimbus Discord isn't like IRC, you don't have to tag people on every response. Keep in mind that every time you tag somebody, they get a notification ping. That can very quickly become annoying and people may block you.
When using Discord's Reply feature it defaults to pinging the person you reply to, click @ ON to @ OFF to stop this - on the right side of the compose bar.
Also... have a bit of patience dude... You're not paying me for dedicated support
ok excuse me
So, try ha supervisor repair
where
SSH or on the console (keyboard + monitor)
and how can i use the console
keyboard + monitor
in what sense
You plug a keyboard and a monitor into the computer running HAOS
Why?
it doesent have a hdmi or display port
What is it?
dusun dsgw-210
And it runs HAOS?
yes i bought the version with haos preinstalled
Time to do a fresh install and restore a backup
now i am trying to do the command with "putty", a ssh im my pc with the ip of the haos, and it gives me this: Got permission denied while trying to connect to the Docker daemon socket at uni x:///var/run/docker.sock: Get "http://%2Fvar%2Frun%2Fdocker.sock/v1.24/container s/hassio_cli/json": dial unix /var/run/docker.sock: connect: permission denied
You can't run Docker commands on the standard SSH add-on
What are you trying to do?
i cant install it
this with the software putty
And a fresh install of HAOS is already unhealthy?
yes
Try an older HAOS version? 12.4 has had issues on some hardware
how can i check my version of haos
supervisor version: Supervisor 2023.01.1
Home assistant: Home Assistant 2023.2.3
thats it
What does it say for the OS?
where is?
Open your Home Assistant instance and show information about your system
That also shows the Installation type
Home Assistant Supervised
what?
so what can i fix it?
i have docker CE 20.10.12, and i need 20.10.17, i read it in the github you sent me
Great, so update
If you don't know how then this is not a method you should be using
i installed homeassistant with an official link
The Supervised method has been recommended against for years
i dont used it i think
https://www.home-assistant.io/installation/linux#install-home-assistant-supervised even tells you not to
You do use that
Home Assistant Supervised
First fix the OS
how
i have docker CE 20.10.12, and i need 20.10.17, i read it in the github you sent me
how can i update the docker
where i need to paste it?
I'll say this again:
If you don't know how then this is not a method you should be using
That has to be done on the console/over SSH
i paste it in ssh
You'll need to be root/use sudo
yes i know
are this 2 command or its one command?
One

