#installation-archived
1 messages · Page 43 of 1
well typically that's what you do
(which I can see it's picking it from DHCP just fine, it's just not sending queries to it! But maybe --servers will override it?)
The problem is that it, when given an unqualified hostname for a machine on the LAN, it is (a) appending a DNS search path which is not correct and does not match the hostname of the machine or the search path configured in resolv.conf and then (b) asking an external DNS server to resolve it, which wouldn't work even without (a).
well alpine doesn't like DNS servers that don't follow spec either
so that throws another 🔧 in
So I tell it that (say) there is an mpd server on oculus, and it can't talk to it because it's trying to resolve oculus.local.hass.io instead and asking the wrong server about it.
im aware how dns works
And I'm not sure which of those is causative, like, is it deciding to ask the wrong server and then choosing the wrong search path based on that? Or is it choosing the wrong search path and then falling back to the wrong server because the right server returns NXDOMAIN?
Either way it's doing something wrong and not respecting the DNS settings handed down from the DHCP server, even though I can see in ha dns info that it's aware of them.
What do you mean by this?
try a portable dns server
Alpine follows the DNS spec and will treat the entire domain as if it can’t be resolved if it receives an error status for either query. Home Assistant will run into many unexpected issues in this situation, particularly around updates and installing software.
Also: ha dns reset; ha dns options --servers dns://192.168.1.1; ha dns restart did nothing. 192.168.1.1 now shows up in both locals (from DHCP) and servers (manually configured) but host is still the 172 address and still what it's using, and there doesn't seem to be any way in HA shell to override that.
Well it's getting an error status because it's appending random garbage to the hostnames I'm giving it! If it used the hostname as written or used the DNS search path from resolv.conf, and used the DNS server configured with --server or the one from DHCP, it would be fine!
i'd still open an issue dude
Like, you seem to be implying that the fault is that my local DNS server is not responding sensibly, but it's not even querying the local DNS at all.
Yeah, I'm halfway through writing one
it won't if it marked an error tho
ur an error
but it was romantic
huh, this is interesting
docker exec -it hassio_dns cat /config/coredns.json shows fallback: true
When did it switch back?
haos: gottem
Yeah, that switched back to fallback: true. Changed it back to false and reloaded it.
im starting to think there is a bug at this point
you're the second person today with similar dns issues actually
I have now figured out, however, that the 172.* IPs are the internal docker IPs -- so that's just HA telling me that it queried its local DNS relay and got NXDOMAIN. Presumably the relay is then querying local DNS (with the garbage search path appended) and getting NXDOMAIN from that, as it should.
ya
well yeah those are docker
So the whole "querying external DNS" thing is a complete red herring, it's all internal, the problem is just the search path
Found it.
docker exec -it homeassistant cat /etc/resolv.conf | grep search => search local.hass.io
This looks similar to https://github.com/home-assistant/operating-system/issues/454, which was marked fixed years ago.
I actually read that post earlier
Filed https://github.com/home-assistant/operating-system/issues/2701 about it.
Aggravatingly, it looks like I can't just edit the resolve.conf inside the container and then bounce the container.
Or rather, editing it works, but next time HA restarts it gets overwritten
Haos probably isn’t for you even without these issues man
Well I'm not about to install supervisor on one of my actual machines
Supervised is never an option
I think he's suggesting a container install
I'm using HAOS because I tried a container install and ran face-first into the requirement to hand-replicate every addon I wanted
It was much easier to set up and administer, but also missing a lot of functionality, and I expected that the "easier to administer" advantage would go away real fast once I was looking at figuring out how to recreate the configurations for whisper, jester, cecscan, etc
Haos should be very simple you’re just one of the few with dns issues
But ya you probably would’ve been done already with docker compose
Yeah, apart from that and the cursed backup archive it has actually been pretty straightforward, and all the issues with the backup were entirely self-inflicted
I have never touched docker-compose and know nothing about it, and I also have no idea how to configure whisper or piper, or configure HA to use them, so that strikes me as extremely optimistic
I prefer haos to keep hass totally separate from my server, network, cameras, etc and run containers on my server not associated with smart home
There’s a pinned comment of mine in #voice-assistants-archived with a ready to go copy paste compose file for whisper and piper tho
Something to consider I guess
in general people who are familiar with things like "resolve.conf" and diagnosing dns issues at the CLI level arent really willing to use haos instead of container installs with full control
pro tip: you only need one compose.yml for all of these services you list here and more. look it up, its easy
Part of the fundamental motivation here is that I want HA to be more of an appliance than an ongoing sysadmin project, insofar as that's possible. I have enough of the latter already.
Anyways, it's super late here, I should have been asleep two hours ago. Thanks for all the help, goodnight!
then you already found the way. haos is intended to deliver this appliance experience 😄
Are you not entertained?!
I used the docker script to start an HA instance. After I rebooted the image is still there but no container? Where did the container go??
I'm pretty new to docker which seems like half the issue.
hmm when I try to remove the image it says
Error: image used by 27ec4c5312eec2cccaeb4e5a36618a6fa6b5455009f10978cb1e21f01ef31a0e: image is in use by a container: consider listing external containers and force-removing image
soo the container is somewhere...
if I do ps -all I now see it. no idea why this is necessary
maybe because it's not running
now that it's running it shows up with ps
🤷
because a container has more states than just existing or not existing
I guess I expected 'container list' to list containers lol
compose is somewhat similar, you get used to it
okay I see, thank you
I have yet to play with compose(or whatever the podman equivalent is)
Hi!
I have two HAOS installations, both running on NUC. On one, I can change the hostname, on the other one this option is unavailable 😵💫
See this screenshot of the two - the one to the right has the Hostname option. Anyone have a clue? The account is admin and has "Advanced mode" enabled.
https://i.imgur.com/X444JTo.png
Fine. Did it with the console 🤓 For reference:
ha host options --hostname <newhostname>.local
(reposting here after posting in #general-archived, hope that's ok)
Hello, I'm trying to upgrade HASS Operating System from 10.3 to 10.5 (running on an Intel NUC, installed ~3 weeks ago), but getting an unknown error, telling me to look at the supervisor log, and this is what I see in there:
23-08-24 10:35:32 ERROR (MainThread) [supervisor.os.manager] Home Assistant Operating System update failed with: Installation error: Failed marking slot kernel.1 non-bootable: grub backend: Failed to run grub-editenv: Child process exited with code 1
I've done some looking but hits seem about old versions and not on HAOS-based setups. Not sure how to follow through here. Could it be some EFI issue? Is there a way to see what error grub-editenv actually outputs?
How can I generate fast-backups of /config without the heavy .sqlite 's?
(besides switching db engine ofc.)
tar?
...But I wanna a handy argument in backup service to just customize content.
I guess I could write a shell_command to do generate it, but a pain-in-the to have to do everything manually
You asked for a way to do a file backup excluding the database. tar is about as straight forward as you can get and is available in the container
ok, I agree that is probably the easiest solution for what i need.
I'm tired and didn't fully test this but maybe something like this
tar cvfz config-$(date -Im).tar.gz --exclude "home-assistant_v2.db*" --exclude "*.tar*" /config
Obviously a good idea to put the backup somewhere else but this should get you started.
Hey guys. I'm installing HA OS in Vmware Workstation on a Windows 10 pc. I've downloaded the correct file from the HA website. I can't get the vm to start because I can't find a way to activate UEFI. The directions seem simple, but I don't see any of what the directions say that I should to activate UEFI. What am I doing wrong?
I don't use this hypervisor but maybe this helps? https://community.home-assistant.io/t/installing-home-assistant-on-vmware-workstation-17/570428
The official docs say
After the VM has been created, go to Settings > Options > Advanced. Under Firmware type select UEFI.
Did this not work?
Thanks anyway, I'll give it a look.
@raven forge, that's exactly what I needed! Thank you!
bump
ok, that task took far less then expected along with automation, tweaking and testing.
And it only costs 3MB to get all of the most important stuff
I'm trying to install the docker version for the first time (because I have docker and don't have a Pi) But the docker run command is telling me that the name is invalid ... and when I put the URL into a curl command, it returns back a JSON string that says UNAUTHORIZED
Sorry, I'm on a different computer typing this.
- Use Docker Compose
- Share the exact command and the exact error
(really hard to debug from vague descriptions)
let me login to discord from the other computer so I can copy paste
Unable to find image 'ghcr.io/home-assistant:stable' locally
docker: Error response from daemon: Head "https://ghcr.io/v2/home-assistant/manifests/stable": name invalid.```
{"errors":[{"code":"UNAUTHORIZED","message":"authentication required"}]}```
What version of Docker? What OS?
Docker version 24.0.2, build cb74dfc on OSX 13.4.1
gah, typo
there's two "home-assistant" in the url
"The host networking driver only works on Linux hosts, and is not supported on Docker Desktop for Mac, Docker Desktop for Windows, or Docker EE for Windows Server."
It's running, there's no reason the image won't run, it just needs a different network option
Gah. my linux server is old and won't run modern docker because of reasons. I know, I need to fix that.
I have a windows 10 machine ... can I run docker under WSL ?
No, but you can run a VM
what about a macvlan network driver?
Still no
I'm getting more and more agitated with Apple these days.
(good.)
Hi there, newb here. I got HA set up within kubernetes behind an nginx ingress. Trying a curl on onboarding.html gives me a reply that seems to look ok, but hitting HA through chrome shows a ERR_CONNECTION_TIMED_OUT. anybody there to help? thanks!
connecting to the wrong thing?
I got it running on the subpath /homeassistant, doing a curl -L my.domain.com/homeassistant/onboarding.html gives some reply, doing the same url in chrome times out. What should be the wrong thing here?
Like, you mean the whole onboarding.html is wrong?
where are you running that curl from? the same machine?
what proxy?
Reverse proxy from ingress controller, it is deployed in a kubernets cluster.
Just checked, curl also works from different machine
http vs https?
curl should tell you what ip and port it connectes to
open the devloper tools in chrome and verify it connect to the same
Hah, just did it from my phone, at least I get the blue onboarding page, but the HA logo is missing
ssl redirection seems to be working there
Have been attempting to configure HA for most of the day. Trying to host it through VirtualBox 7.0.8 on a Windows 11 machine. I know not the best choice, but wanted to toy around before I got in deep. Everything seems fine until I get to the "Prepping Home Assistant" page and it will hang for an hour at a time. I've scrubbed the forums and discord before finally resorting to asking here. Probably not looking in the right spots, because I'm still not sure what I'm missing. Seems every post wants to blame DNS but I've run many DNS commands with zero success. Currently I'm stuck in a loop between updating, downloading, and the following error:
[supervisor.docker.interface] Can't install ghcr.io/home-assistant/qemux86-64-homeassistant:2023.8.4: 500 Server Error for http+docker://localhost/v1.42/images/create?tag=2023.8.4&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)")
23-08-24 16:27:08 WARNING (MainThread)
Any assistance or pointers to the correct location would be appreciated!
dont think its dns, it times out making a outbound http connection
firewall? networking configured wrong?
IPoAC?
haiku time?
It's not DNS.
It can't be DNS.
It was DNS.
I can't even make this up. Somehow I figured it out 20 minutes after shamefully asking in here. It was DNS.
it sucks being right all the time
what exactly was wrong?
prob no nameserver for the vm or something
Well first, my Googling abilities. But it never crossed my mind to check the DNS of my host. The Windows 11 machine was using a DNS provided from my ISP, once I set it to 1.1.1.1, the HA installation immediately finished.
pretty sure that would give you another error, more along the lines of not beeing able to resolve the name
eww isp dns
your router isnt handing out something else via dhcp dns?
and no pihole/adguard setup?!
interessting, what ip does it resolve ghcr.io to? just out of curriousity, i wonder why they would deliberately return wrong ip addresses? some form of blocking?
No, nothing set up yet. I've been playing with pihole, proxmox, HA now. But I haven't gotten a foundation for my home lab. Kind a just toying with it all until I feel more confident.
To be fair - that wasn't too hard with DNS 😆
just jump in
you can use tteck scripts to get going and then redo it yourself to learn if you want
https://www.derekseaman.com/2023/06/home-assistant-proxmox-ve-8-0-quick-start-guide.html this is a pretty recent guide if you want some hand-holding
https://tteck.github.io/Proxmox/ scripts im talking about - they're popular here
Thank you! I'll definitely be checking these resources out!
that's why it's an easy side to pick
I've just installed home assistant and I can't access the website, http://homeassistant.local:8123. I don't have your conventional ethernet connection, the ethernet plugs into my pc and is told to send the wifi network, and it works on other devices. On my raspberry pi 3 it says up the top no ethernet found or whatever. Any suggestions?
I suggest you buy a switch and a few ethernet cables because that sounds like a terrible environment to run servers (anything, really) in and your PC has to stay on?
How did you install HA? How did you connect your pi to the network? Up the top?
well what the OP really needs is a router by the sounds of it
|| router I hardly know 'er ||
Has anybody else experienced this stuck onboarding page with an empty logo? Funny thing is that also home-assistant.log is empty - where do I even start debugging here?
Just want to feedback on the very much appreciated help you gave the other day. Reducing CPU allocation from two to one really made a more stabel HA. Thank you 👍🏻 🙂
I got it running on the subpath
the pi is connected via an ethernet cable that runs to my pc, which was setup to send wifi signal through the cable. up the top of the console log.
yeah i need one, where we live we cant though. the only thing we could get is starlink. I have a portable hotspot thing (like a phone just only for wifi connections)
I suspect your WiFi to ethernet bridge isn't working but I'm not familiar sith that. Have you tested that?
yeah like i said it works with other devices
It wasn't really evident that you tested other devices with the same cable.
Can you show the message on the pi? If you run HAOS, can you also show the output of network info?
sorry.
It says "net: no ethernet found"
also i know the wifi is working cause i can access the observer page
So can you share pictures of the command and such?
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.
I've taken the liberty to upload the image you PMd me here so others can chime in: https://i.imgur.com/bhutM9c.jpeg
It looks like the network cable is indeed connected but it didn't receive a DHCP address. I assume the other devices were configured via DHCP as well?
are you aware you can get a router you can put the SIM card in ?
Is there a known issue with alexapy where I can find some docs on dependencies?
apparently the Alexa Media integration requires alexapy>=1.26.8 but with a dependency on aiohttp>=3.8.4 ...but home assistant is only compatible with aiohttp==3.8.1
ok, i upgraded aiohttp to 3.8.5 and home assistant did not complain that time, and alexapy to 1.26.9
restarted and Alexa Media throws ERROR (MainThread) [homeassistant.setup] Setup failed for custom integration alexa_media: Requirements for alexa_media not found: ['alexapy==1.26.8'] in the log
ok, different question: how do I update my HA version for an ubuntu install?
What kind of install?
I have had it on a Pi for almost a year now, and haven't really had to touch it until now
Home Assistant 2022.12.5
Frontend 20221213.0 - latest
What kind of install?
Core, Container, Supervised?
Well if you are running Ubuntu, it won't be Supervised
Where do I look to find what I need so I can appropriately answer your question? Not seeing any references to that
it's in a /srv/homeassistant/ directory?

Maybe this helps: https://www.home-assistant.io/installation/#compare-installation-methods
You can also go here and make a screenshot of what you see: https://my.home-assistant.io/redirect/info/
It's running from my Pi, headless, so I dont' even really need to mess with it
ok, I made a backup.. says Core 2022.12.5
If you run docker ps on the console - does it show sth like Home Assistant?
is this my doc then?
https://www.home-assistant.io/common-tasks/core/
command not found, for docker
yes
ok, taking the Core reference from the name that it gave to my backup created in the UI, and trying the directions in the documentation for Core updates.. it seems to be working
why isn't that in the "about" section?!
Before you try to update, you should read the release notes for 2023.1 to 2023.8 - there have been breaking changes, which might hurt you. Especially the Python requirements
oooh.. thank you for that warning!
Some exceptionally shitty ISPs have their nameservers configured so that instead of NXDOMAIN it tries to send you to a search page or the ISP's tech support page or a cutesey "looks like you made a typo, here are some ideas of what you might have meant" page.
any chance you have instructions for pointed HA at a different python install?
can I accomplish that with /srv/homeassistant/pyvenv.cfg ?
hello every body
Hello hello
i have a problem when i install home assistant
i Install Home Assistant Supervised on debian 12
after i excute command: apt install
apparmor
jq
wget
curl
udisks2
libglib2.0-bin
network-manager
dbus
lsb-release
systemd-journal-remote
systemd-resolved -y
i excute command: curl -fsSL get.docker.com | sh
and i get an error:
“curl -fsSL get.docker.com | sh” returns “curl: (6) Could not resolve host: get.docker.com”
please show me how to solve that error
Search ‘supervised’ here to see why you should not use that install method
Either use container or haos in a vm/bare metal
thank you very much
ya it's a trick
i will install HAOS on X86-64 computer
it really just shouldnt even be an "option"
https://www.ebay.com/itm/385670237223 link i've shared previously for an $80 mini pc "m910q" there are many variants on ebay for reasonable prices with 8gb ram or higher and 128gb ssd/nvme or bigger - this is a great option if you need cheap hardware
used laptops work well too cuz built in power backup with the battery
cool
This is a guide for 3.10 - needs to be changed accordingly for 3.11.
https://community.home-assistant.io/t/upgrading-homeassistant-core-in-a-python-ve/420027
pyenv is handy
Great to hear it!
you can TOTALLY plug a starlink into a router. It's actually better that way. Starlink throttles your speed when you connect more than I think 4 devices to the base station. If you plug it into the router, it think it has one device, and you can hook as much to the router as the connection can handle.
You'll have to order the ethernet adapter from starlink if you have one of the newer units.
I have three people right now with starlink and unifi networking and they're very happy with it
I'm pulling my hair out with regards to kernel out of memory errors and my pi4 frozen. Analyzed docker stats and swap and free memory and glances can't see any problems. Happens often and whenever I do an update.
Any clues to help please ?
Possible that my nvme drive is corrupt and it's actually a disk issue? Is there a way to run fsck?
How much RAM, and what install method?
It's been running for years. Latest hasio. 4 gig ram. Pi 4
What add-ons?
Got rid of myriadb still an issue
Running paradox alarm interface but I don't think it's that
Memory usage is very low
What add-ons?
Can I image link the add-ons?
Yes we'll memory free is always around 3.3gig
Add-ons. Excl integrations are.
Esphome. Advanced ssh and web terminal. File editor. Let's encrypt mostly off. Mosquito broker. Paradox alarm interface.
And more, lots more based off the image you posted
how do I remove an entry under settings > devices & services > integrations ?
Oh well I only installed glances and influx and sqlite web today. So can't be that
i want to completely remove one and start over
#integrations-archived can help with integrations
What must I do I can't live like this anymore
I use home assistant all day
I'll pay money
I'll buy hardware
I'll do anything
😂
Well... answering questions and not leaving stuff out because you've decided it's not relevant... that'd help a lot
after like 7 hours, of setting up a new machine on my network, and building a proxmox server, and HA install, and figuring out the restore takes a while.
I listed everything besides for things I installed today to try fix this. This has been going on prior to today.
I DID IT and i feel sooooo good about not using virtualbox anymore
@grand pivot thank you for answering me
I've also increased swap space utilizing those init strings in the ssh plug in. It's been like that for a while. Bad idea ?
Not sure if u know what I'm referring to?
Tinkerer do u have any suggestions please ?
Moar ram
That and chase the cause of the OOM, but you've picked what you share, rather than telling us all... so I'm out
**I am hoping someone can help. **I just installed a fresh copy of Home Assitant VMWare VM and restored from a full back. I've saved a few full backups since then.
Now when I restore from backups, after its done, everything looks the same, nothing was restored.
I tried restoring from my original backup, same thing. Before I go again and install another fresh copy of Home Assistant VMWare VM and restore from backup again, I need to know what to do in order to prevent that from happening again
I did not. I told you every addon that was installed during the crashes. The photo today was after I installed additional addons that couldn't be related to the issue as it was crashing before that. I tried to explain clearly you don't understand ?
I dont see any errors in the log related to the restore
It's not difficult to understand is it?
Well, have fun
What can I do to try find the cause of the problem ? Please help me
I'm not having fun I have a problem
stop some of your addons. at least 32098 of them
I have so few running. Influxdb added a few by itself such as nginx and kapacitord and cronograf. But the issue was prior to that. I only have advanced ssh. File editor. Let's encrypt. Mosquito broker. Paradox alarm interface.
It's really not a lot which is odd. And I have usually around 3.4gb memory free of the 4gb.
When running out of memory, adding more stuff cannot be a bad idea
What HAOS version are you running?
Yeah that's what a have
Yes that is correct the other devices are getting assigned dhcp addresses
I have, running a different os on there works fine, the only one I have trouble with is ha
My suggestion stands.
If you configure it with static IP, does it work? That would point to dhcp problem.
otherwise, what hardware is your deviec? (I don't remember). Is the NIC anything exotic?
Your screenshot shows HA is detecting the device, it's plugged in, and has a mac. Sure looks like a dhcp problem to me
PS: it's slightly unusual for your device to still be named eth0. Normally they get renamed to some annoying device-specific name like enp0s18
(but I don't have a rpi with haos for example)
People in this forums ingested I add influx and glances to try determine what is causing the issue.
Yeah, me. If you want to see historical memory per-process you simply need some additional tools.
I suggested having top open in the past while triggering an update so you can see live what happens: <#general-archived message>
I have no recollection of you getting back to me with any results. My memory isn't the best though.
I did an update with top open as well as glances. Didn't see anything spike at all. All I can think of that might be causing the issue is the paradox alarm interface addon.
And it did crash on the update yesterday
You could downgrade and do it again with top open sorted by memory usage
Is there a way to limit the memory on a specific add on so only that add on can crash
Technically yes but I'm not aware that HA exposed that functionality anywhere
Oh OK good idea. How can I downgrade. ?
I have a feeling it crashed actually on first boot restart after the upgrade. By thelat time obviously top wasn't running and logs are gone
OK got your message on how to up down grade. Great
See my link.
You can use docker update to set limits on running containers but it's manual and I don't recommend that: https://docs.docker.com/engine/reference/commandline/update/
I also have previously manually adapted my swap memory on the ssh addon
Do u think that could acyltually cause an issue
The default is 1gig. I've got it to 4
I don't know
are you talking about this? https://community.home-assistant.io/t/how-to-increase-the-swap-file-size-on-home-assistant-os/272226/184
haos 10.x changed the way it handles memory or something like that and people had issues showing maxed out ram if they were using that workaround linked above
Yea like that. However I'm not swapping I have 100 percent of it free. It's never used it at all
But still crashes with out of memory errors
And I still have 70 percent of my ram free
#general-archived message here's a link to falstaff talking about it a little bit a while back
personally i'd open an issue on github if there isnt one already in the haos repo
Interesting.
I'll try work out if it's a plug in doing this first.
It's quite difficult to work it out though especially if you reliant on them to be running most of the time and this issue isn't happening every day
So latest is I disabled add-ons that could be causing the issue and I stopped them from start on boot.
I downgraded to 2023.8.2. System booted I saw it was on this downgraded version. Then crashed. Didn't start anything manually.
Question for all...running home assistant on an rpi 4b off an SSD. Upgraded the OS last night and no longer seems to be booting properly (can't connect and the rpi is no longer connecting to the network.) I assume there is no recourse beyond doing a clean install and installing a backup. Is that correct?
I'd connect a monitor and keyboard and see what it displays
That's the plan. Just have to wait for the kids to give me some time to do it.
When I used a pi4 sdd boot I had no issues with usb power only. It really depends on which ssd and sata to usb adapter you use
I can only tell that powered ones usually work without any issues while chances using non powered ones are significant that it won't work
I did pick up a couple before I noticed
Star tech 3.0/3.1 work well
I setup home assistant on a yellow, no issues with install but as soon as I had it up and running my Nest WiFi stared to have all sorts of issues, mesh points going offline, etc. I unplugged the yellow and the wifi stabilized. I did not enable any nest wifi (or other nest/google) integrations. Has anyone seen this behavior before?
#yellow-archived might have
Thanks, I’ll repost there
For what it's worth, it seems like one of the USB ports on my rpi went bad. Switched ports and we're back in business.
Nice
so, i'm trying to install on my macbook m1 chip (used a old dev release for virtualbox) and it didn't work. i can't get past the post-booting part.
and it aborts itself every time
i actually followed a new tutorial, and got it fairly working. still installing debian for the new tutorial
@scenic grove there's UTM as an option for M1.
Utm?
yeah, a tutorial had that. following that
UTM, a VM software made for mac m1, m2 i think?
Is that like parallels?
I got a HAOD VM running as a test and it worked
I’m not a mac os guy just played
yep
Me either. There was something I had to do to get networking going but not in front of it now to check
no, free and actually better i'd say
i'm at the debian almost done installing stage
Another victim of the Supervised install trap?
?
i'm done with dependencies
but
mv /etc/network/interfaces /etc/network/interfaces.bak isn't working
nor
curl -fsSL get.docker.com | sh
can't resolve host
so, thought about it and took a break, gonna get a raspberry pi as you can use the software installer/flasher and boom
A Mini/(U)SFF PC might be a better choice.
If you want help with your issue you need to elaborate on isn't working as that's not really helpful
https://ibb.co/Wvckyry
Hi all I'm still struggling with these OOM errors.
Swap is never being used at all and memory usage never exceeds 20%
It seems like the kernel OOM killer shouldn't be doing this at all. I've disabled most integrations and add-ons. Still doing this after updates on first boot.
Each docker container is using extremely small amounts of memory. It's so strange this. Any ideas please
In general it's best to show this information rather than telling us vaguely what it is.
issue
Hey everyone, thinking about starting from scratch with a new install but I dont want to have to link all my devices again to my zigbee coordinator (mostly light bulbs) is there something I could copy from a previous install that might help?
#zigbee-archived could help there - if you tell them which Zigbee approach you're using
Hello I have an home assistant app to control my cameras and gates for my properties, I want to give access to only my gates so the tenants can use the home assistant app to open and close them, but do they have to create and install home assistant to and go through all the installation guises or can I just add a user
This is more of a #general-archived question
But the creating and installing ha is only for a brand new instance. But note that creating accounts and giving people access they could view all devices added to your instance so cameras and gates
You would be better off creating a new instance just for gates and giving people access to that, though again everyone will have access to everything
There are no permissions as such in HA
Ha is designed as something for a group of trusted people (eg a household) not for strangers to access it
do any of you guys know of any programs that can be used like that
wait so your telling me if I'm the admin and I make a sub group that has access to only control the gates, i cant assign a user to only that sub group
Nope. Pretty much every home automation system is designed for homes.
There is no such sub group type to create
lol, so much for "assistant"
Average landchad be like
It's the word "Home" that is important here.
It is an excellent assistant for a home and household. Not so much for hotels, shared access buildings, etc. nothing I've encountered is designed for that
yes but say i dont want my kids to control certain items in the house, shouldn't i be able to not give them access.
You don't give them access to HA then.
I've shared a limited subset of my devices with HomeKit. But that is also designed for households.
but what if I still want them to access certain items like the garage
You'd need to trust them.
There are some solutions like exposing select entities to Google Home or using Telegram bot commands that you can implement in that situation but HA itself doesn't do user permissions and it's not designed for implementing access control.
i just did a reinstall of my HA server, on to an ssd using usb boot. It's restarted a handful of times, but i have had the home assistant is installing page could take up to 20 mins. i've never had it sit this long is my install borked? or keep waiting?
nevermind i figured out how to see client output.. its borked.
i think i was using a power supply that wasnt powerful enough and it effed it up. dont be me boys.
With a Pi? If you're using an SSD and booting a Pi from USB it's recommended, probably more like required actually, to use a powered USB hub.
If your install freezes, it might be an issue with setup not reaching DNS service. I had to force e.g. OpenDNS via DHCP server to HA server during installation and only then it went forward. After that, restoring backup took something like 35 - 40 min.
What did it show?
Just force homeassistant IP and DNS in your router configuration.
DNS server
And - actually, reinstalling didn't work from USB boot, so eventually I flashed the HASS image using USB-C cable. Even on Mac it was not a hard thing to do.
But there is a plan to add permissions. Wonder what the state of this is.
https://developers.home-assistant.io/docs/auth_permissions/
I've seen no signs of that being actively worked on
fresh install of proxmox 8.0.5 & HA running in VM, added the usb passtrough already for the conbee2 ( and updated ) but no one on forums seem to find dev serial by id. How do I find the path to setup a zigbee integration?
There, at that link
yeah ive looked there.
If there's nothing there then it hasn't been passed to the VM
My dream is LDAP support and permissions like that for HA
Since you're an expert tinkerer, Do you happen to know if anyone is working on either of those or if it's just dead right now
how do I migrate my homeassistant container onto a new system?
i use docker compose
My dream is LDAP support and permissions
Restore your backup to the new system?
You are backing up already?
So... just restore your backup, or rsync the config folder over
Can someone point me to a working troubleshooter for missing add-ons section? Seems like it happens a lot to newbs like me lol
Usually when you either installed HA via a method that doesn't have addons, or you're following an ancient guide that doesn't show that it moved to Settings
It can only be the 1st one maybe? I didn't know some versions didn't have addons?
I have Home Assistant 2023.7.3 Frontend 20230705.1 - latest
add-ons are docker containers and require the supervisor managed docker ecosystem (not a plain docker container install). So HA OS or the "supervised" install (which you should avoid)
Brilliant. So I must have core then?
What install method did you use? Docker?
Open your Home Assistant instance and show your Home Assistant version information
Ah, they moved that info
Open your Home Assistant instance and show information about your system
That one
@open ferry I converted your message into a file since it's above 15 lines :+1:
Ah yeah I divined that from the link. core's the only one that doesn't have it. I know how to pick em....
So I guess I have to look into the supervised installation method. Was wondering when that was gonna come up
It looks like Truenas supports virtualization, you could run HAOS in a VM.
That was my first attempt. But trying somehow messed up virtualization and now none of the VMs start
You don't want to do that. Supervised has specific requirements that you're probably not going to be able to maintain on Truenas without your HA breaking.
Should I just bite the bullet and break out a retired thin client?
add-ons aren't required to use Home Assistant. Are there any in particular you were thinking you would need?
Well HACS
that's not an add-on
#hardware-archived message this was shared yesterday lol
Cuz apparently I need it to integrate my xiaomi devices
HACS works with all methods
The guides say I need ssh to install it
Ah.. haha well I have that
HA OS doesn't have a shell without making use of add-ons, so that's why that is there
Well finally a break
I'll probably still have to move to dedicated hw. Is it better to do it now before I've configured too much?
A backup of your main config directory including all hidden directories would be copied over to the new installation, so it's not that big of a deal
Ok, cool. That gives me a bit of room to experiment
Thanks for the help. You and the others. It's a bit overwhelming all of the options
Mini pc++
Something like that. I have a couple old Lenovo M92s I can raid
Perfect
Yeah those Lenovo Thinkcentres are great machines for running HA, they get recommended on a daily basis over in #hardware-archived
Glad to hear that!
HACS installed. That was relatively painless once on the right path
Curious is this stream error normal? It's a stock install
Setup failed for stream: Requirements for stream not found: ['ha-av==10.1.0'].
Setup failed for default_config: (DependencyError(...), 'Could not setup dependencies: stream')
You need Python 3.11(.4), and there's some other issues around ha-av on Core - related to the version of FFMPEG
Ah that is another issue I've been trying to solve. How to upgrade it to get rid of the nag.
I'm stuck on 3.10
You'd rebuild the venv on 3.11, however you do that on TrueNAS
Not sure what that means, but I found that I have both 3.9 and 3.10 versions installed after installing HAOS (this is a jailed env)
Simply running the python3 installer ignores both.
Well, you only get HAOS with a VM
Ah ok wrong terminology. Core
You need to install Python 3.11 in the jail
Sounds so simple... and yet .... lol
I'm sure it's my lack of experience w python tho. It's probably super simple
Well, pyenv is the way, but ... may not be easy with a jail
My problem may be that I was holding out for a package installer. Might need to compile it myself
OoOoOoo pyenv did it
Maybe not. It says it installed 3.11 but HAC knows nothing about it.
You have to create a new venv, activate it, install HA in there, and get the jail to start that new version
On Linux this is easy
So I have to make sure I'm running python 3.11 before installing HA?
I see. Probably pull it out and do that tonight then., thx
hi, i joined here when i started with ha today cuz i knew i would encounter something i cant fix on my own;
So i set up my random pi 2 to run ha, it worked as intended, it even auto detected my wled setup; my problem is that i wanted to plug it into the second router i have and i cant access it; i tryed searching on the internet and some places mentioned port forwarding and such and i got scared so im looking for help here;
basically the setup is that internet comes in from the provider, that is the main modem/router (kaon from vodafone), and i have a second one (dlink) connected to it via cable, and my pi is connected to that second router right now. again, the pi setup worked while it was connected to the main router;
how can i find that out (networking is literally my nemesis, i know very little about it)
Well, did you connect the WAN port of the DLink to a LAN port on the Koan?
from kaon lan port to dlink "internet"; the dlink has 5 ports, one yellow input labelled "internet" and 4 numbered output ports
Then those are two separate networks
Why did you do that?
What's the point of the DLink?
originally wifi didnt cover the whole house, and later when we switched provider 2 out of the 3 tvs got connected to that one cuz it was closer
Then you want to:
- Disable DHCP
- Connect the LAN port instead
Then it'll be a switch and access point, not a router
That'll solve your new problem, and others too
i suppose i do that in the dlink's panel, but problem is i cant get there; there is no ip written on the thing just "dlinkrouter.local" which times out, and from the kaon's panel i can see an ip but that times out too
Can you tell us its model number?
it is a dlink dir-842
Apparently it uses 192.168.0.1 by default. See page 3 here: https://support.dlink.com/resource/PRODUCTS/DIR-842/REVC/DIR-842_REVC_MANUAL_v3.00_WW_EN.pdf
ERR_CONNECTION_TIMED_OUT
in the kaon panel it says its under 192.168.1.3 but that times out too
How is your device (where you opened the browser on) connected to it?
my pc is connected to the kaon router, my pc is where im trying to connect
i just format my homeasistant machine, and wanted to start from zero, now i open my system and i constantly get login atempt. checking from ip and its from my pc. how can i understant whats it
Can you check the local ip of your computer?
So the ip range is different and the other router would most likely not let your pc connect anyway due to firewall. Not sure if fiddling with static routes helps here
So basically you need to be connected to the same router to connect with each other
can i do that from an android device trough wifi? cuz i cant really connect my pc to the dlink directly
No I meant you need to physically connect into the same device. Your PC's cable and the PI's cable go into the vodafone box, for example
ah
ye sry my brain skipped a step i meant to say that to acces the dlink panel
The dlink might be able to be set up to only act as AP and switch but I'm not sure
Only if you're connected to it. You should get a 192.168.0.xxx address
I think i broke something, i turned off the option "dhcp server" and now the panel times out
Assign yourself a static ip
But as tinkerer said, it might be best to use (it as) a switch
Anyone familiar with the python install of Core on linux? I'm at the venv activation and it's trying to do stuff to folders that don't exist
What stuff and what folders?
Actually I think I know where things went sideways. I'm having to 'adjust' the instructions for FreeBSD, and I may have missed a dependency ...
I guess my question is, if I'm installign this in a jail, is the venv stuff still necessary?
Hola alguien puede explicarme como puedo actualizar la version de home assistant? yo tengo instalado un CASAOS y dentro de el la version de home assistant docker compose 2023.6.3
In English please
Sorry!, can someone explain to me how I can update the version of home assistant? I have a CASAOS installed and within it the version of home assistant docker compose 2023.6.3
Here is the guide for updating the container version (there Docker Compose can be selected). Before updating, you should check the release notes (for 2023.7 and 2023.8) for breaking changes.
https://www.home-assistant.io/common-tasks/container/
Thank you!. I have followed the steps and I already have the updated version. The problem now is that the installation is clean. With nothing of what I had before. How can I recover everything I had? I also have a backup that I made from home assistant type xxxxxxx.tar
You shouldn't have lost anything. Whatever directory you mapped for your config directory should still be there
Unless you didn't map anything...
And to be able to recover what I had, how can I do?
Well, did you map a config directory?
yes
i put this:
docker run -d
--name homeassistant
--restart=unless-stopped
--privileged
-e TZ=MY_TIME_ZONE
-v /DATA/AppData/homeassistant/config
--network=host
ghcr.io/home-assistant/home-assistant:stable
Then you shouldn't have lost anything
That's not valid
You're missing the :/config on that line
Extract the contents of the tar file into that directory, fix that line, bring the container back up
ok i try it thanks!
Hello!, I have extracted the contents of the tar file and put it in the /config directory, but I start the home assistant and the clean installation follows me without the data from before and with the updated version. What am I doing wrong?
Did you update your run command?
-v /DATA/AppData/homeassistant/config:/config \
And is that where you extracted the tar to?
I have not run the command yet, I just extracted the file in the same directory that is /DATA/AppData/homeassistant/config and then restart the home assistant to see if everything is as before, but the installation is still clean in the updated version.
Well... then you still haven't mapped that folder so of course your files aren't there
Hi, a small problem on my linux. i need HA and plex via docker, but both bind port 1900/udp for DNLA => the 2nd service that starts (plex) complains because port already in use.
any idea on how to manage this please ?
Was setting up ha on rpi4 with ssd and suddenly everything became unavailable - according to the logs. when connecting the pi to a monitor there's a constant spam of 'unable to read ...' or 'failed to rotate'
failed to rotate seems linked to logrotate. check disk space or logrotate config (and file rights)
disk space should be fine, brand new 240gb ssd. i cant ssh or access pi via command line - would it be fine to pull the power and plug it back in? Last time this happened on an sd card, thought switching to ssd would fix 😦
never use rasp, you could still restart it, not a big task 🙂
yeah rebooting the pi fixed it, just want to prevent this from happening again. disk space according to ha is 98% free. checking the file system though /dev/root has 0 available space, thinking maybe ha tries to modify files in there and that's where the read only problems come from
The SSD, is it connected to a powered USB hub?
Because if not that's probably the source off your problem
ahhhh yeah that's probably it. too much data transfer over the ports on pi? i just have a usb to sata cable
Just not enough power on the USB ports
There's a post from somebody suffering from that just about every other day
🤣 yeah can imagine, must be tiring having people like me asking about it everyday. thanks for the reply, i'll pick up a powered usb hub
The problem is that it's not an obvious problem, and the official Pi docs don't even hint at it
yeah that and a lot of tutorials out there don't tell you about powered usb hubs and just use usb-sata cables or even tell you to just use an sd card 💀 so long my last sd card lmfao
Hello, just installed and got a few things up and running using latest version. However, I am not seeing the Configuration option in my side bar even though I have enabled Advanced in my profile. I also restarted the system after enabling. Any suggestions? Thanks
@vast fjord you're referencing something old and out of date. Do you see Settings?
Settings > Dashboards > energy
Excellent. Thanks
Hey folks, I've used HA for quite a long time, but abandoned my home lab over a year ago, didn't have the mental energy to fix it.
I am looking to completely rebuild my homelab, motivation behind it is to learn ansible.
Since AWX wants to be deployed into kubernetes (but docker also works):
Did someone manage to get Home assistant working with a "full" K8s setup?
Still also new with K8s, too.
I saw a few people trying to run HA on K8s, but I heard there are a lot of problems.
I kinda dont want 2 container hosts...
Assuming you're aware that not all integrations will work, HA works there
Anything that uses UPnP/discovery/mDNS/broadcasts won't work
I mean I can get it to work 😄
But probably wont worth the hassle... how are you running it right now @humble mirage ?
Not in K8s 😛
For those that don't want a dedicated host/VM (aka HAOS) then native Docker make most sense
I have no problem running HA in a VM, my base for my home lab server gonna be proxmox anyways.
Maybe I'll go with HA as VM and then the rest in K8s/k3s w/e 😄
Speaking of weird and overkill things to do, did someone alrdy try to use apache kafka instead of mosquitto :D?
Kafka does MQTT?
oups, it doesnt. I thought I read somewhere that Kafka can also "speak MQTT", but doesnt make that much sense tbh 😄
I can't access the local home assistant page on my phone or my newly flashed ns panel but it works just fine over my laptop any idea why?
could be a number of things
are they on the same network?
are they using hostname instead of IP?
My phone says connection refused and on my ns panel name not resolved
Yes my home wifi network
Hi, i am trying to update my home assistant from core-2021.6.6 to core-2023.8.4 (and supervisor from 2021.06.3 tot 2023.08.1) but the system is "unhealthy" and cannot connect)
😂
just got my rpi, and it came with a home assistant sd card, so.. need to either format the sd or solve it by updates
well
formatting the SD card and installing a current version
gonna be way less painful
straight to jail
Still not working
Webpage not available
The webpage at
http://homeassistant.local:8123/auth/authorize? response_type=code&client_id=https://home- assistant.io/android&redirect_uri=homeassistant://auth- callback could not be loaded because:
net::ERR_NAME_NOT_RESOLVED
dns, use IP or fix your dns
Where are the dns settings?
Do I have to use the terminal for that?
dns settings are handled by something else on your network
usually gateway or router, or something else
Connect over the IPv4 address?
yes
Connection refused
how can i acess my config file remotely?
Trhough SSH, Samba, Or the UI w/ any addon that does that
# Load frontend themes from the themes folder themes: include_dir_merge_named themes
frontend:
automation: include automations.yaml
script: include scripts.yaml
scene: include scenes.yaml```
this is my configuration.yaml
no
clear your browser cache then, other than that, not sure.
the first two lines look wrong there
and it's !include, not include
Hi! I watched this video from CrossTalk Solutions
and I installed Home Assistant in the docker desktop application for Windows. I don't have a smart home nor a segregated IoT network nor a server. I need help with setting up separate networks for my personal network, guest network, IoT network and maybe an IDIoT network to identify the IoT devices. Also, I'd appreciate being pinged so I'll for sure answer to you.
Well, HA isn't supported in Windows like that... so you're off to a bad start
You would however need to put HA on the "IoT network" to enable it to talk to the devices there
https://www.home-assistant.io/installation/windows shows only two options, neither are Docker
I download the windows installer for docker from https://docker.com and then I installed the HA image in Docker Desktop application. This is an image of the app.
https://www.home-assistant.io/installation/windows shows only two options, neither are Docker
AKA ... not supported
Plus:
You would however need to put HA on the "IoT network" to enable it to talk to the devices there
(though, most people use VLANs without understanding networking, and just get themselves in a mess)
you will have a bad time trying to use HA in docker on windows, as windows doesn't support the host networking mode, so HA won't be able to discover devices.
I've got to go now. I'll refer back to here in the morning for me. (It's 12 Am for me.)
I'm struggling to complete the VBOX setup on macOS 12
The VM aborts immediately.
I get several lines of error codes in the logs, mostly relating to the display settings.
Happy to share the whole or partial log
This is one of the messages I get in the notification center of VBOX:
VM Name: HomeAssistant
hv_vm_create() failed: 0xfae94001 (VERR_NEM_INIT_FAILED).
Driverless mode (VERR_SUP_DRIVERLESS).
Result Code:
NS_ERROR_FAILURE (0X80004005)
Component:
ConsoleWrap
Interface:
IConsole {6ac83d89-6ee7-4e33-8ae6-b257b2e81be8}
I've reinstalled VBOX
Well, could I instead run HA in VirtualBox? As I previously said I don't have a server but I may in the near future tell me parents to buy me a raspberry pi.
Also, what about WSL? Is that a sustainable environment for HA core? If yes, does WSL use the host's resources?
Yes WSL does, it has to
Core isn't recommended, but if you're comfortable with command lines it works
Is there a difference between HA and HA core?
To clarify, are you replying to me?
So it both has CLI and GUI in WSL?
Oh yes. I forgot about the web UI 😅
Is a raspberry pi a could installation method or is WSL better or worse?
Pi with an SSD on a powered USB hub is better than WSL
A VM is better than WSL
However, a second hand small form factor PC will likely cost the same as the Pi while performing better and being more stable
For now as I don't have any smart devices set up yet and available installation methods for me are:
- Docker
- WSL
- VirtualBox VM
which would be best?
On Windows we've already explained why (1) isn't an option
As for WSL vs VM that also has been answered
ok. thanks.
Rather than a pi get your parents to buy you a $100-150 used sff pc. Way more powerful and useful
What's a sff?
Ah. ok.
Like a Lenovo thinkcentre or the dell equivalent
Yeah got it.
Also, if I install HA on a VM in VirtualBox I don't have an Ethernet connection with my router (well it's actually a gateway (according to Bing) )
Please keep in mind that the bridged adapter only functions over a hardwired Ethernet connection. Using Wi-Fi on your VirtualBox host is unsupported.
Your main machine is wifi only?
In my house there is a desktop with the gateway right next to the door to my house and my PC is in my room a couple meters away from the gateway
Run an ethernet cable to the gateway? Otherwise wifi is probably ok to play with, if you only access it from your pc
It's gonna be "blocking" entrances and exits to rooms. There is hall that connects my living room with a kitchen and another smaller hall that connects my parent's bedroom with mine and the bathroom.
WiFi is my only option for now....
For WiFi I use a TP Link adapter. Is that a problem?
Virtual box abstracts that away, so doesn’t matter
ok
If you’re going to be there for a while maybe get a cable run properly in the walls?
But if you get a sff box you can leave it next to the gateway and wired in anyway
For me WiFi is the only option for now because I'm gonna somewhat soon I'm gonna move my PC do the apartment I've got in the ground floor which is mainly used for tutors and I'm gonna leave my PC in the bedroom. It's a work-in-progress that apartment 'cause it only has 1 desk for studying and 1 bedroom with no beds and only some closets cupboards and drawers.
Fair enoug
I just set up HA in VirtualBox and then started it. It prompted me with this screen and then tried Slot A failed and tried again and again.
What kind of failed? I don’t know anything about windows btw
Can’t watch it. On phone
It's ok.
This video has nothing to do with HA!?
I installed HA in VirtualBox according to these instructions.
It shows the process of the VM trying to boot up.
Maybe check your link again. It shows you starting InsaneCraft
Sorry abt that.
Wrong vid 😅 🤦
This one should be the write one: https://www.mediafire.com/file/90e2en0do5e9e3u/Desktop_2023.08.30_-_11.09.51.01.mp4/
I have no clue why this happens but you could try spamming the down arrow while booting so you can select another slot. Maybe something else happens.
Otherwise maybe share your .vbox file and/or try re-downloading the vdi and re-creating the VM with it. It also says OK=0
Ok. I'll try to change the Slot
Slot B shows
error:unknown filesystem
The rescue shell for Slot A has the same output as the normal slot A
But um.. 2nd try with the slot a rescue shell and I think it's booting up.
And it's booting
yep
And it's up and running nice.
Docker image not launching on Synology NAS - could anyone help?
I have Docker running, downloaded the Home Assistant image from the registry, download to my Synology is complete, but nothing happens after i have hit Launch. Done it number of times, uninstalled Docker and all over again, but no cure unfortunately.
I think there was an issue with the docker hub url for HA on synology.
#installation-archived message
I'll try to link using the URL, thx. Strange that it did download the latest 2GB image...
unfortunately did not help. again a download, but nothing starts after hitting launch
anyone experience with irresponsive images in Docker on Synology? have also restarted my NAS.
I'd try to start it through CLI (command-line) if possible
sorry CL? - newbie here
Command line. Is there anything in the logs?
is there any help how to do that - sorry newbe here
the log just mentions it downloaded the image
Add image from docker.io/homeassistant/home-assistant:latest
That does not look right. Did you follow the instructions here? https://www.home-assistant.io/installation/alternative/#install-home-assistant-container
yes, i did, followed it all - strange right?
simply not starting - and unfortanately not familiar enough with CLI
Because the docs state stable, not latest, for example. I don't own a synology device so I can't really test it myself
I previously had homeassistant running via python in a virtual environment on MacOS with https://www.nordicsemi.com/Products/Development-hardware/nrf52840-dongle - it was autodetected by home assistant without an issue and added Thread/other stuff nicely.
I'm now trying to setup an Ubuntu server installation of home assistant on a laptop with the same dongle, and while ubuntu and home assistant in docker setup ok via command line, all that nice stuff the nrf52840 usb dongle added is not being detected. lsusb shows the device is definitely being recognised as it shows up in the list, but I'm so new to ubuntu that I have no idea how to get from A to B here.
Apologies if there's something I've missed in this, just not sure why it worked in the first method but not the second, and not sure if it lies in something I did wrong in ubuntu or something I need to do with home assistant
You have to give the device to docker run/compose. I suggest the latter:
- https://docs.docker.com/engine/reference/commandline/run/#device
- https://docs.docker.com/compose/compose-file/compose-file-v3/#devices
What doeslsusb -tandls -l /dev/serial/by-id/say?
If it doesn't show up as serial you can also checkls -l /dev/for something likeUSB/ACM/TTYor just share the output
thanks a lot! Guess i have to give it some more time, by first finding out how to CLI on my nas
The post you replied to was actually meant for Greg 😄
ahhh
tried a CLI via Task Scheduler I found from MariusHosting on how to install HA on Synology. Gives me this output:
Current status: 126 (Interrupted)
Standard output/error:
docker: Got permission denied while trying to connect to the Docker daemon socket at unix:///var/run/docker.sock: Post "http://%2Fvar%2Frun%2Fdocker.sock/v1.24/containers/create?name=home_assistant": dial unix /var/run/docker.sock: connect: permission denied.
does this ring a bell?
The user running that might have to be added to the docker group
i am running as admin
syno utilizes sudo
I still had to add it manually. Even create a docker group
Will 2gb ram be able to run a couple dockers or even the ha os itself?
Temporary while I get a low cost laptop or some such
Yes but I haven't seen them officially called dockers. They are containers, not a shoe 😉
It's not really the amount but rather the software that is running inside the container(s). VS Code server usually uses more memory than mosquitto, for example.
As long as it's not a pi trying to run it 😄
@thorny vector I converted your message into a file since it's above 15 lines :+1:
@raven forge that was in response to my above query, apologies, the bot converted it!
While I'd use the /dev/serial/ path that looks correct. I believe you don't even need this with privileged. I'm not familiar with the integrations for this but you probably have to give them the same path. Not sure how their detection works
Hmmmmm I might just try a Skyconnect 😅 this is driving me crazy
thank you for your help though @raven forge - in the meantime if I wanted to learn a bit more about Ubuntu Server so i understand the basics for when that arrives, how would you advise going about it? no worries if you're not sure, just thought i'd ask!
I actually couldn't tell you just one source. I also don't use ubuntu and it's been a while since I read a getting started with linux guide.
The digitalocean articles are usually good. The debian manual or its admin handbook, linuxjourney perhaps.
There's also guided linux courses on coursera and the like if that's more your thing: https://www.classcentral.com/search?q=linux&free=true
If I have a problem/idea I get my info from different sources and then usually test in a VM. I reference things most of the times so I can't tell you about the coherence of big guides.
My suggestion would be to get some virtualization software, create a VM, create a snapshot and break things while trying to achieve what you want to do whioe following guides.
Restore the snapshot and repeat until you're happy with the result. Then do it for real. I use VMs this way multiple times each day when helping people here.
Hey not really an HA question, but is there an easy way to resize partition on a raspberry pi on ssd ? I try with an usb stick but it wont boot 😭
Ehh, having /home on a separate partition can be nice for reinstalls at least
its a folder bro
copy the folder
there should be gparted isos out there for aarch64. that should do fine on a pi ssd,
except its haos - in that case better not do that
i think gparted also comes with raspberry os, so just flash that to a sd and boot it
Anyone care to help?
Running home assistance on linux port 8123, getting error "site cannot be reached"
We need more details
How are you running it
What IP are you typing in the browser, etc
Right now it's the same as ringing a mechanic and saying my car doesn't work
No details is useless
Start by checking the home assistant logs
and the problem is a windshield wiper
Hey, I need some help. I'm trying to run home assistant in a KVM but I'm having some network issues. I can't seem to get a connection to my local network through a bridge from the VM to the network interface. Can anyone help?
How did you create the VM? How does the bridge config look like?
I created the VM using virt-install
if you didn't specify a networking type it's NAT by default
I eventually ended up going to the system hosting it itself and opened a desktop interface and after not getting anywhere I installed virt-manager so I could visually see everything and set everything. I ended up making the bridge in virt-manager which is called virbr0
Somehow now there's two virbr0 bridges in network connections...
is NetworkManager enabled?
It looks like it.
disable it
What does bridge link or brtcl show say?
Virbr0, the id of it and the stp enabled status which is yes. No interfaces
@raven forge
It should probably have an interface. Maybe follow something like this to create the bridge: https://computingforgeeks.com/how-to-create-and-configure-bridge-networking-for-kvm-in-linux/
I was just told to disable the networking. That's why.
Let me re-enable it and see what it says.
Well now there's nothing. And it won't even pop up again. I think I have to reboot the system because the interface has to reinitialize
Unless there's a way I can get the interface to initialize again after being disabled
How did you disable it?
Yes but how?
I didn't say disable networking, just the network manager
Sudo systemctl disable NetworkManager
That's what I meant
That does not necessarily stop it though
that just stops it starting up
Try the same command but with enable to reverse it
Done
status to check its status, stop to stop and so on
What do the bridge commands from earlier show now? You can also use systemctl enable/disable --now btw to also start/stop it
I see another bridge labeled BR0. And the one that previously had labeled virbr0 They both don't have interfaces.
I'm suspecting that's the issue
Because in the KVM earlier, it couldn't resolve the IP
That's weird though because I thought I assigned them to the main network interface
Wait now networking is off?
And now I can't re-enable networking.
Okay, I just logged back in and it says I am connected to auto ethernet
I'll try getting the bridge to work now
Virtual machine connection details say that it is forwarding to virbr0 so I'm assuming that isn't set up correctly because it's forwarding to Nat on there
but you want bridging not NAT
I have a couple options. Nat, routed, open, isolated, and SR-IOV pool in virtual machine manager
Then it tells me what interface I would like to use. I'm going to use the main ethernet interface
Now I'm back to where I was before, I am looking at the virtual machine and I type the command. Ha Network info. I see the first section that says docker and it gives me an IP address starting with 172. Under that it says the interface is hassio it says host internet false, interfaces connected, true enabled, true interface ENP0S2 which is I think the one I want there's no IPv4 address gateway is no, there's no name servers and the method is set to auto
Why not just share a screenshot?
Hi all, I'm on Supervisor 2023.08.1 and trying to upgrade to current 8.4, however something goes south. The web-interface never comes back up on port 8123. I can ping the 10.x.x.x/8 IP address just fine and I can reach the HA console directly. As my CLI kung-fu is no the strongest, any tips how to troubleshoot this? (Should be said I'm running it on a Hyper-V VM with one NIC configured as bridged to the local network with it's own address. Should be said it's been working fine so far. Also have a working pre-upgrade snapshot, so I'm not dead in the water per se, just can't upgrade) Any help appreciated - Thanks!!
10.x.x.x/8 ? where'd you get that? Microsoft?
that's 16,777,214 usable IP addresses....
That is a well known internal class-A network, like 192.168.0.x is a class C and 172.16.x.x is a B class. Point is it's a valid IP range and no, Microsoft has nothing to do with that 🙂
Run network info in the VM's prompt. Also check supervisor logs
no one uses a /8 in a home network, so no it's not a 'well known' subnet to use
Classful addressing is so '80s
Hey Impact, thanks for actually answering my question, contrary to the two other (now blocked) useless jackasses. I'll have a go and see what HA's yelling about when I upgrade again
well that escalated quickly
I'm curious now, I wonder how many others are using a larger than a /24 for a home network
everyone using docker does
well that's on one host typically
what do you mean, people have more than 1 host?
im talking about for the entire lan
Even /16 is excessive for a large smart home setup
But, many folks are stuck with classful in the brain so would jump from /24 to /16 bypassing all the other options
/32 for point to point links is fun
ummmmmm
if you hate the person just say that
/32 is one IP... I think you mean /30
Nope
I use /24 for each of my (3) vlans, so that's ~768 addresses, which should be fine
And non-octet boundary subnets make my brain hurt :0
ur an octet
I do use a /16 because numbers need to be ordered
So, problems with NFS storage and HaOS......
problem of HA being launched before the NFS mounts are ready
meaning that HA reports them as failed, and storage starts to fill up the local docker images
You can hit "Update" on the storage config to mount them manually after the reboot, but then you mount over a directory that's already been written to, and it gets ugly to clean it up
ie unmount it manually, copy stuff out, mount it manually, then copy stuff to the storage path
i've not seen that happen at all with my mounts
can you share some logs?
ignore the message from this and use the link
Sadly we're not mind readers (any more anyway, not after the last time we tried). Please share the YAML and any errors so we can see what you've done.
well, it doesn't hit homeassistant.log
supervisor?
yeah - I'm not quite sure where to grab that from lol
is there a linky that shares that wisdom?
Right now, the install is:
OS Version: Home Assistant OS 10.5
Home Assistant Core: 2023.8.4
that's just homeassistant.log?
OH! I never noticed the dropdown menu at the top right hahahah
There's nothing in the supervisor log that mentions NFS at all
nothing in the host log either - but I wonder if I'd have to reboot it - like maybe its gone out of a log buffer?
possible
Do you use IPv6 for NFS or IPv4?
I mean, I wonder if its a race condition between activating IPv6 and NFS
let me reboot the entire system and see if I get anything in the logs
oh, I can't paste the image of the failure lol
@lofty shale I converted your message into a file since it's above 15 lines :+1:
I'm just trying it again with IPv4 addresses instead of IPv6 for the NFS target.......
Yep - it works fine with IPv4 NFS mounts - no errors
but with IPv6 NFS mounts, it fails - until you hit the Reload button on the error notification.
Hi 👋
Trying to install Home Assistant on Ventura MacOS 13.4.1. I downloaded “haos_ova-10.5.vdi” and when I am trying to install it I see an error message saying “There is no application set to open the document “haos_ova-10.5.vdi”.” Can someone help please ?
Well vdi is a virtual disk. I presume you wanted to run HAOS in which case you need a program to run the virtual machine
I have downloaded HAOS from https://www.home-assistant.io/installation/macos
any recommendation ??
And did you read the rest of the site?
You want virtualbox or install HAOS bare metal on something else
Yes I did, Wanted to try with HAOS but if that doesn't work (solution not found) i can go with virtualbox
....
You are going with virtualbox. That's what's going to be running in virtual box, home assistant OS
so i click on virtualbox it downloaded haos_ova-10.5.vdi and that i can not install
Because you need virtualbox installed as a program. .vdi is a virtual box virtual drive
Its like downloading a .docx and wondering why it won't install word
Ohhh i See 😲
Thanks 🫡
Hello everyone!!! Nobody tried to install on Mac OS M1?
I think some have successfully set up a HAOS VM using UTM on Mac M1s
I'm having failures running container 2023.8.4 on aarch64. 2023.7 worked fine, but 2023.8.4 is in arestart loop much like: https://github.com/home-assistant/core/issues/86589 ```/hass$ podman exec -it homeassistant /bin/bash
odroidc2:~$
odroidc2:~$ python
Python 3.11.4 (main, Jul 23 2023, 12:40:43) [GCC 12.2.1 20220924] on linux
Type "help", "copyright", "credits" or "license" for more information.
import jwt
Bus error
odroidc2:~$ ```
I've just rolled back to .7, but it smells bad 😐
Addons are missing from a fresh docker install, how do I fix this?
root@router-master:~# ifconfig br20
br20: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 192.168.20.242 netmask 255.255.254.0 broadcast 192.168.21.255
no /8 just for localhost 🤫
you don't get addons in a docker install
you install your own containers and get the same thing
all addons are, are docker containers 😉
So where do I report the bug that IPv6 NFS shares don't mount properly at HaOS startup, but IPv4 ones do?
Thanks
that would be github
tbh not sure exactly what that is an issue with, I think it might be supervisor
perhaps someone smarter than me could chime in
but also:
Supervised system is not really maintained right now, also the mounts feature is more an OS feature as on supervised, you can do it at your own
are you running HAOS?
yeah
then that's not to do with you
that's for supervised installs, that person is on debian 12
I did also read that post 😉
righteo 🙂
its just a bit hard sometimes on what's kind of an applicance device to know what slots where
Ok, filed this: https://github.com/home-assistant/supervisor/issues/4517
There's no error messages in any of the logs to be useful, so hopefully the explanation is pretty straight forward.
one thing I found strange - there seems to be no builtin thing to restart the nginx addon when the LetsEncrypt addon updates the cert
so I built a dodgy NodeRed automation lol
Loads the old cert into a string, starts the LetsEncrypt addon, waits 1 minute, then reads the cert into a string, if they differ, restart the nginx addon lol
Yeah - I would have thought that the LE addon would have a way to restart the nginx addon if it updates the cert, but nope lol
I have synology nas, and I'm hosting homeassistant there via docker, any tutorials on how to connect it to my esp32?
What firmware is the esp32 running?
latest, v3.2.0.0
🤔
Does that firmware have a name?
v3.2.0.0 ESP32-WROOM-32-AT-V3.2.0.0
Bruh
What are you trying to do, where did you buy it, what integration are you trying to use?
People flash all kinds of firmware like wled, esphome, etc to an esp32. What you’ve quoted means absolutely nothing
To do what lol
Bluetooth proxy?
We can help you real gud if you give us information to help you with
I want to configure it to connect to my wifi, so i can connect it to my home assistant docker
😢
to do what
Stuff and thangz
I want to do automations, once a trigger is called it should push some voltage to a desiared pin etc
You'll want to flash a firmware like ESPHome or Tasmota then
https://esphome.github.io/esp-web-tools/
I've installed it with this
I'm following up there, I'll update on how it's going
thanks
Hello, does anybody knows how i install my rf 433 in hassio?, i have https://www.nodo-shop.nl/nl/aanbevolen/127-rflink-arduino-antenne-.html
Hass.io was the old name for the appliance like install option, that uses Docker. It was renamed in January 2020 and is now called Home Assistant OS.
So I've installed and it's working well, but when I try to connect to my home-assistant I get:
Can't connect to ESP. Please make sure your YAML file contains an 'api:' line.
I need to edit the .ymal but I don't know how since I can't connect in the first place...
Most people install esphome addon then create a new device and flash firmware over usb. Or plug in usb to your pc and use the web flasher then add it to hass then OTA update it with more yaml changes
"...then add it to hass then OTA update it with more yaml changes"
This is where I lost you 
(I'm using docker, not the supervised)
then install the esphome container or just use the webflasher instead
mstone linked it above
Wasn't there an option to keep history data for longer than 30 days
No
If you want granular long-term stats for your entities you should use something like VictoriaMetrics/InfluxDB and Grafana
For long term history a database like InfluxDB is designed for
uh how would i do this
i set mine to a year to see what happens. currently my only issue is backups and updates are taking a really, really long time
like is there a ha built in feature to just do it for me or do i need to extract that data on my own
There's addons for the mentioned software. Then read the addons' docs and: https://www.home-assistant.io/integrations/influxdb/
How can I change my Host Name from the cli? I've got a 2nd instance of HA on an old laptop that only has wifi, no ethernet, and I don't know how to enable that wifi adapter yet from HA, but I do want to make sure my production HA does not get confused when i do get this 2nd instance online.
ha host options --hostname newhostname.local
I've ordered a usb to lan adapter, as I like to run it wired anyway, but is it possible to activate the wifi?
But I'm not sure, if those setting survive a reboot. Never tried it myself
But when you get access to the UI with this method, you can go to Settings - System - Network and set the Wifi stuff there. That should be persistent
You're a walking almanac 🙂
Just good in finding stuff 🤣
I think I should exclude the .local from the newhostname.local, as the home assistant url now shows as http://testha.local.local:8123
It's really local now 😆
The instructions to activate wifi from cli are perfect. Thank you!
