#installation-archived
1 messages · Page 33 of 1
you could try plugging the wyse in by ethernet to your pc and run a quick and dirty DHCP/DNS server
that's easy
Idk how to do that but I'll try it
https://giphy.com/gifs/DxUiFqLgDVC00 @frank perch
do I need to set anything up or just plug it in and let it run
well you'll have to just do connection sharing prolly
but you can edit things like your host file on your PC
but something (I'm betting Alpine) doesn't like your DNS server on your router
so we are going to bypass that
HAOS should add support for USB cell phone tethering
Smart you wanna let's make a fr?
lol
alright finished moving everything, ready to power on
I started out as a hardware guy
then I ended up being dragged kicking and screaming to software
you got the connection sharing figured out?
probably not
ya, nope, didnt do that right
and this is just for testing, this is not supported for production, nor is to be considered legal advice, or to be taken seriously, with laces in your shoes
my cat's about to break my door down, brb
so I dont have the dropdown in step 8 if that matters
and when I click that box in 7 and hit ok it knocks out the wifi connection 🥴
are you on windows 10 or 11
10, I have one of the class action amd cpu that win11 doesnt like
I could also try with my laptop, it is also win10 tho
This is so frustrating
it's just one of those things man
ya, and I know it will be worth it in the end
you'll get it
its just one of those, I spend all week getting railed by shitty programming to come home and get my ass kicked on personal programming problems
but ya, I cant tether this. When I tell it to share it drops the wifi connection so there isnt anything to tether
might be doing it backwards?
do I hit share on the wifi adapter?
I was doing it on the ethernet connection
idk, lets find out
worst thing is it doesnt work
Right-click the adapter with the internet connection you want to share on Windows 10 and select the Properties option.
so that would be your wifi
since youre on wifi for internet, you wanna share that to the ethernet
if you shared the ethernet, it would create a wifi AP with your wifi which is why you got knocked off I think
well I had that backwards in my head
shouldn't have to right away
let's see what this does
since we are in the land of experimentation now
the waiting for supervisor to startup sure takes a while
Depends on the hardware
You can tail the supervisor logs to see what's happening
wondered why I couldnt see it, forgot it got a new ip
been a long day
supervisor_internet: true
what logs would I check
23-05-27 21:56:41 INFO (MainThread) [supervisor.host.services] Updating service information
23-05-27 21:56:41 INFO (MainThread) [supervisor.host.network] Updating local network information
23-05-27 21:56:41 INFO (MainThread) [supervisor.host.sound] Updating PulseAudio information
23-05-27 21:56:41 INFO (MainThread) [supervisor.host.manager] Host information reload completed
23-05-27 21:56:41 INFO (MainThread) [supervisor.resolution.evaluate] System evaluation complete
23-05-27 21:56:41 INFO (MainThread) [supervisor.resolution.fixup] Starting system autofix at state CoreState.RUNNING
23-05-27 21:56:41 INFO (MainThread) [supervisor.resolution.fixup] System autofix complete
23-05-27 22:02:45 ERROR (SyncWorker_0) [supervisor.docker.interface] Can't install ghcr.io/home-assistant/generic-x86-64-homeassistant:2023.5.4: 500 Server Error for http+docker://localhost/v1.42/images/create?tag=2023.5.4&fromImage=ghcr.io%2Fhome-assistant%2Fgeneric-x86-64-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)")
23-05-27 22:02:45 WARNING (MainThread) [supervisor.homeassistant.core] Error on Home Assistant installation. Retry in 30sec
23-05-27 22:03:15 INFO (SyncWorker_1) [supervisor.docker.interface] Updating image ghcr.io/home-assistant/generic-x86-64-homeassistant:landingpage to ghcr.io/home-assistant/generic-x86-64-homeassistant:2023.5.4
23-05-27 22:03:15 INFO (SyncWorker_1) [supervisor.docker.interface] Downloading docker image ghcr.io/home-assistant/generic-x86-64-homeassistant with tag 2023.5.4.```
docker logs -f hassio_supervisor, journalctl -fu hassos-supervisor.service
To view older ones use it without the f/-f
It's what's used for the DNS_SERVER_FAILED issue type but I don't see it here
says there isnt a -f flag
You used login again yeah?
Remember, the prompt has to look like # . not ha >
CTRL + C
desktop browser do what? sorry
ty
They want to you to open: http://_checkdns.home-assistant.io/ in your browser
Interesting I get an error on that
Error 1034
Ray ID: 7ce2f8f32d261435 • 2023-05-28 02:12:24 UTC
Edge IP Restricted
What happened?
You've requested a page on a website that is part of the Cloudflare network. The host (_checkdns.home-assistant.io) resolved to an IP address that the owner of the website does not have access to.
Afaict it's only used to check if it can be resolved and it can
I dunno, normally an error means something isn't working...
i think he's saying it's just supposed to be a txt record not resolvable to a site
nvm derp that doesn't make sense
https://github.com/search?q=org%3Ahome-assistant _checkdns.home-assistant.io&type=code
https://github.com/home-assistant/supervisor/blob/334bcf48fb7e1572933c211ca6b8fef46de0deb6/supervisor/resolution/checks/dns_server.py#L62
I only learned about this URL yesterday when I researched the DNS_SERVER_FAILED thingy.
well based on the notes on that link it points to dns server (netgear router) being fucked
idk what I'm looking at
well _checkdns.home-assistant.io resolves to 1.1.1.1
for me at least
so I guess if that's all it's used for
Supervisor does lots of checks, this is just one of them. If it fails it should emit a DNS_SERVER_FAILED like yesterday with the broken 10.0.0.38 DNS ip.
I don't see if this time but we haven't seen all the logs yet
He had the same issue after changing it to authoritative servers like CF as well I think
@valid plover you tried nmcli setting dns to 9.9.9.9 and 1.1.1.1 right?
As well as your gateway (netgear ip)
Also at this point you’ve moved off the windows pc and back to the router right?
cmdline: git clone -v --recursive --depth=1 --shallow-submodules -- https://github.com/hassio-addons/repository /data/addons/git/a0d7b954
stderr: 'Cloning into '/data/addons/git/a0d7b954'...
fatal: unable to access 'https://github.com/hassio-addons/repository/': Could not resolve host: github.com
'.
23-05-27 20:25:41 ERROR (MainThread) [supervisor.store] Can't retrieve data from https://github.com/hassio-addons/repository due to
Seems bad
I'm grasping at straws
I just nmcli set those, do I have to do a save command or anything
Idk I copy paste nmcli shit
Kill it with fire
I'm clueless and need all the help I can at this point
wait, could I try to use my phone to wifi hotspot to the desktop and tether the thing to that?
Well in your defense you chose the hardest hardware to install on + have what seems to be a router that sucks at doing dns
it didnt seem that bad from guides I read, and Pi's are expensive af these days
I went to rebuy what I have running octoprint behind me and its jumped crazy
That thin client would work fine running octoprint I think
if its dns it doesnt matter if I install it on a toaster
I know
that would be neat tho
I have five printers with five pi’s but nowadays I’d buy a mini pc and virtualize multiple instances I think
fair enough
Anyways I think you just have to put the interface down then up but maybe you do need a reboot after changing nmcli stuff
I just rebooted it
also running desktop on phone hotspot
YOU HAVE GOT TO BE KIDDING ME
THAT WORKED!?
HOLY SHIT
great stuff
now... here's the other issue
it will need updates and to install addons on it's own
ya, but now its 100% confirmed the router
So this just means, your router needs a swift kick in its arse
yep
Well, at least we now know through rigorous testing that it's narrowed down
That was an awful way to pop your cherry
no spit or anything
Can you post some logs from your failures on home assistant OS GitHub? Like open a new issue if you can’t find an open one and link to the closed issue I posted above
Points to netgear routers being the issue
good idea
This is why I ditched their trash firmware literally on first boot of the router
Great hardware but the firmware was written by drunken rabid monkeys
Eh
‘Great hardware’ from a consumer grade all in one piece of shit lol
The netgear Nighthawks specifically
Well the x10 does the what... 3rd fastest routing I think out of the routers tested
It's a war on open source
What is
Well DDWRT is on mine
I haven’t touched ddwrt since the wrt54g days
I've only run DDWRT since rhe 54g days
I literally refuse to buy routers that don't support DDWRT
I like my UniFi silver simplicity for home
Cuz usually they are trash
I’d run opnsense if I wanted life on hardmode tho
I used to run a hardware router myself and gave up on that over 15 years ago
I have an ASA on my desk and a PIX in the closet if you really want hard mode
I do not
Well time to sell it then
What
You have the r6900P?
r6900v2
That's supported
I didnt see it in the list
If you dig around on that site you can prolly find a newer build
But that's what came up for me when I searched
No sweat
You’ll have to reconfigure it and rejoin all Wi-Fi devices if that isn’t obvious btw
Could use the same SSID and password but the BSSID might be different. Worth a try
Hi, I'm running an old version of HA (2021.6.6) in docker using the official HA container and I want to upgrade to the latest version, how should I proceed?
@
You update it the same way you update any of your docker containers
Except because it's so old expect a lot of breaking changes
Did you use docker compose or docker run
Hello all, i want to run HA container on a rock4C+, but i have some difficulties. I follow steps from: https://www.home-assistant.io/installation/linux#install-home-assistant-container but got stuck on the libseccomp requirement and the script (which folder to use for config files). Any advice is much appriciated! Thx, P
Hey all, does HA require any open ports?
What do you mean?
Well the page helped me a ton yesterday trying to install HA for the first time and we came to the conclusion that my nighthawk router was causing all the headaches. We ended getting it installed by tethering my wyse to my desktop and having the desktop connect to my phone mobile hotspot and the install went fine
we set the dns servers, but I was not sure if maybe a port forwarding of some form would help
It was dns issues which apparently netgear routers cant do properly
Port forwarding is only needed for remote access
If you're doing port forwarding for local access then things are all kinds of fucked up
netgear + github was no bueno
I dont have any enabled, just curious
I'm in uncharted waters
Port forwarding is only ever for accessing things from outside a network
whats annoying though is my desktop connects to it just fine
You either port forward or use a tunnel like cloudflare or tailscale if you want to expose it yourself out wan
TIL
I thought you were gonna install ddwrt to see if that unfucks it
Cuz you won’t be able to run updates, install addons, etc if it still can’t reach GitHub
Its another rabbit hole I'm reading into and not wanting to jump down unless necessary lol
That’s like saying you’ll replace your tires after you sign up for and drive in a race
Have fun with the first corner 
there's not too many things to be "afraid" of concerning DD-WRT, if you can setup a stock router, you can get DD working
however it will allow you to do some advanced things, when you are ready
but the DNS server in DD-WRT follows the DNS spec, and works.....
the only thing that I 'lost' when I flashed my router was WAN MAC cloning
Hello, I need some help with a network issue on a fresh install on a 4gb RPi 4.
I'm trying to use HA over wifi.
Problem is, I can set it up once from the UI and it will work just dine, until I first reboot. After I do, HA disables WIFI and it then becomes impossible to turn it back on.
Every time I do try to turn it back on, HA just barfs a " 'Activating connection failed: Check network settings' occurred X times" error message. I've checked the core logs, but there's nothing there except the same error message. Didn't realize to check supervisor logs.
Only way to "fix" it, is to do a fresh HA install, which will work once and then do the same as above.
Everything in the UI is set to auto (DHCP), for both IPV4 and IPV6 and nothing on the network changes between reboots
What do the supervisor logs say?
@paper kraken I converted your message into a file since it's above 15 lines :+1:
Hey folks – struggling with this error: Building wheel for miniaudio (pyproject.toml) did not run successfully. I installed following this guide https://home-assistant-china.github.io/docs/installation/armbian/ and a virtual env
That is a forked project not official and certainly not recommended.
ah so I can ignore? Also getting messages like self._raise_for_failed_requirements(name, missing)
File "/srv/homeassistant/lib/python3.9/site-packages/homeassistant/requirements.py", line 270, in _raise_for_failed_requirements
raise RequirementsNotFound(integration, [req])
homeassistant.requirements.RequirementsNotFound: Requirements for apple_tv not found: ['pyatv==0.10.3'].
sorry, I'm a bit lost... this is the first time I'm trying to work with HA on a raspberry pi alternative machine
look at the domain you linked
The link you posted above literally has china in the name
the version you have is likely very very old
I'm trying to run HA on this https://www.amazon.com/dp/B0BQG668P6?psc=1&ref=ppx_yo2ov_dt_b_product_details
does not matter
And armbian isn’t supported
So was directed to this article https://home-assistant-china.github.io/docs/installation/armbian/
Anywhere
from the HA subreddit
tell them they are wrong
Then that person needs to be taken out back and shot
OK, so what would you recommend I do? Is it still possible to run HA with that machine?
Look up a guide to install it in docker if you can’t do it without
here was the post I made on the subreddit https://www.reddit.com/r/homeassistant/comments/13st1je/comment/jlvz9q3/?context=3
OK
Btw, a 3 day update from my problems about constant crashes on 2023.5.4, downgrading to version 2023.5.3 did fix my issues, no crashes since. Not qualified enough to know if its a virtualbox on windows issue, or maybe a processor issue, or what, but after a few attempts at downgrading from a clean install of 2023.5.4 to 2023.5.3, there was one that was succesful without the OS crashing and restarting and had 0 problems since. Probably a combination of new OS + hardware, but again, Im not an expert.
This was on a clean (more or less, some drivers and other basics to run most software) windows install with latest virtual box, latest windows 10 + all updates, i7-3770k processor (a bit old), 8gb ram, 2tb ssd, rtx 2060 6gb gddr6 graphics, with a wifi network card (which i know isnt officially supported but dont see why that would be an issue on a specific version of the OS). Virtual box configured with 2 processors + 2048mb ram (though I tried raising to 3 processors + 4gb ram and still had the same crashes)
(also tried reinstalling the entire vm from scratch at least 5-6 times, with the exact same issues before downgrading, nothing worked)
If needed, I can create a new VM to reproduce and post logs.
Hi,
im trying to install HA by proxmox, and im unable to do apt-upgrade it fails with this code: Err:1 http://deb.debian.org/debian bullseye InRelease
Temporary failure resolving 'deb.debian.org
Anyone know what i can do to make it work. ?
intel nuc i3
Hm i setup with my dns..
any directions i should look ?(its the first time for me ) 🙂
First time for what?
Can your proxmox host reach the Internet? Can it resolve domains?
hm ping: google.com: Temporary failure in name resolution
I guess you used wrong networking information in the installer. How does your /etc/network/interfaces and /etc/resolv.conf look like?
Thank you, somehow it changed my dns, got updates now 😛
Now im shell install HA, and it ask for q35 vs i440fx machine
Both should work. I usually use the default i440fx unless I have a reason not to. Also see: https://forum.proxmox.com/threads/q35-vs-i440fx.112147/#post-483840
Ok,
disk cache Default / write trouh?
Stick with the defaults for now
Okay, maybe not for everything. Here I'd recommend host for best performance
OK.
CPU cores ?
At least 2, more (overcommiting to a point) doesn't usually hurt either. You can always change it later
my specs says 2 (Threads: 4)
Someone should have selected "Default" settings
i got 16gb ram, how much should i allocate for HA vs Proxmox ?
Depends on the add-ons you're going to run
Yes deffently.. but dident wan to use 4gb ram only..
Somewhere between 2 and 8
Thank you. Now it seems to have been install. My ssd is 128gb but VM says 32gb ? Can i update that ?
You can increase it but I would hold off until you need to. HAOS should automatically grow the fs if you do. 32 is just how the image was created
32 GB is also the minimum, if you're going to stick with it you don't want to keep too much entity history
yeah an i do run a lot of history and collect some data.
Do you know how to change the size ?
Hardware tab, click the disk and at the top there's a disk actions button. It's better to store history in VictoriaMetric/influx. More efficient
I don't have hard numbers and depends on data but you can easily store billions of things with less than a gig of storage
I use VictoriaMetrics. Not a fan of Influx's syntax. It's still compatible with the influx integration
Okay sounds like im gonna look into that..
Thank you for all the help so far really appriciated!
I recommend starting with the basics and gradually progressing by utilizing resources such as Google to search, read, learn, and adapt as your requirements evolve.
deffently - ive been using rp for the past years, and installing on nuc was completly new to me. and the guides i found wasent the best ones.. So im happy that we have a comunity like this to help out 🙂
Plenty of good user submitted guides can be found here https://github.com/tteck/Proxmox/blob/main/USER_SUBMITTED_GUIDES.md
hey :)
Is there a way to add a root CA certificate to Home assistant OS store ? or at least to the Home assistant container on it ?
I 'm having issue with a couple of integration that don't allow to bypass SSL, and it seam the general resolution is to "remove ssl" because 'it's local anyway"... Not a fan of this solution 😅
I made a fix to add my cert to the python certifi list, but it look like this isn't enough for the Fullykiosk integration
If you need it for another thing, see https://community.home-assistant.io/t/let-home-assistant-trust-a-personal-certificate-authority/184917/22
my supervisor seems to think my HA is not running whereas it is -> [supervisor.auth] Home Assistant not running, checking cache / Anyone has an idea how I can dig deeper into this ?
seems related to another message i get -> Can't execute restart while a task is in progress
when I try ha core restart
if you have ssh setup, you can try to check docker status
Or power it off fully and try to turn it back on
there is a lot of failsafe on hassos
I'm trying to set up Home Assistant Core through WSL https://www.home-assistant.io/installation/windows/
I keep getting stuck on "Setup of bluetooth is taking over 10 seconds" when I get to the "hass" part
Warning (MainThread) [homeassistant.setup] Setup of bluetooth is taking over 10 seconds.
It should eventually continue
I left it for an hour and nothing
I have a fast computer so idk why that's happening
There’s an open issue on it
To be honest... Core on WSL isn't a great choice
right but it should at least function
Once the bug is fixed
There’s an open issue on it 
have the link for it?
I’m on mobile sorry
👋 I have a home assistant installation working fine, using docker, and not using host networking, but the server can't be autodiscovered by companion apps (though manual connection works fine).
Is there an additional port that needs to be opened/accessible for that to work correctly?
Can’t he technically manually setup ssdp and other things if he hates himself?
It'll be using mDNS/uPNP, which requires broadcasts, which don't work without host mode
You can theoretically setup a reflector in the docker net and another one on your lan but you shoot yourself in the foot by doing that
Thanks kindly, host mode it is
Hi, I'm stuck getting my Z-Wave controller to work again after a re-install using a qemu virtual machine on Debian. When I try to start the Z-Wave JS add-on, I keep getting a RangeError: Index out of range (https://pastebin.com/SYDNXtad). I don't know how to debug this. Can someone point me in the right direction?
jump to #zwave-archived
thanx, will do
Heey, So im missing the last thing on my install.
But Proxmox / HA install dosent see the GPU so i can enable HWACC.
does anyone know how i can make Frigate addon to see the GPU ?
Hey guys, im trying to shrink the .ova deployment in esxi, i managed to shrink the vmdk to 16gb which is what i want, but now it wont boot.
is there a way to get the homeassisant ova to work with 16gb instead of 32gb of storage?
32gb is the min recommended
the computer im using does not have enough storage for 32gb as im running over vms
HA doesnt use that much storage
Hi, I just installed home assistant to a VM on windows (Virtualbox), but I can't access the url page homeassistant.local:8123
has anyone talked about/played with the idea of running homeassistant in a highly available fashion? say, with two yellows and a carp vip or the like between them?
or a primary/failover model? I'm reticent to just assume that the one device will work forever without ever having any problems.. an figured that if one wants to depend on a thing, having a standby thing is a good way to shore up that possible expectation miss
starting from no failure plan at all, one's time is better spent getting automatic off-machine backups, and once you have automatic off-machine backups, the cost-benefit curve gets very unfortunate. and zigbee/zwave are especially hard to failover non-manually
Hi I just installed home assistant to a
Morning, I just installed HA using docker compose. At first glance it is running fine, but doesn't seem to have access outside of my local Raspberry PI so when I tried to install HACS I got an error wget: bad address 'get.hacs.xyz'. looking into the logs I see Logger: pychromecast.socket_client
Source: /usr/local/lib/python3.10/site-packages/pychromecast/socket_client.py:398
First occurred: 10:39:30 (1 occurrences)
Last logged: 10:39:30
[AFTR-56(192.168.0.56):8010] Failed to connect to service ServiceInfo(type='mdns', data='AFTR-56._googlecast._tcp.local.'), retrying in 5.0s
If I ping google.com from the PI it works fine, if I use docker -exec -it homeassistant bash and then try to ping google.com I get a bad address
the pychromecast error has nothing to do with your hacs issue I think
you probably have a host issue when docker cannot resolve dns
thanks, the host appears fine? networking within docker appears to be the issue, ie this fails but works from the host outside of docker
[root@Plugbox homeassistant]# docker exec -it homeassistant bash
Plugbox:/config# ping google.com
ping: bad address 'google.com'
[root@Plugbox homeassistant]# ping google.com
PING google.com (142.250.179.238) 56(84) bytes of data.
64 bytes from lhr25s31-in-f14.1e100.net (142.250.179.238): icmp_seq=1 ttl=115 time=19.5 ms
64 bytes from lhr25s31-in-f14.1e100.net (142.250.179.238): icmp_seq=2 ttl=115 time=17.9 ms
looks like a generic docker issue, how did you install ? on what ?
installed on raspberry PI, using a YML file. I. have another docker service that works fine
I used this compose file
version: '3'
services:
homeassistant:
container_name: homeassistant
image: "ghcr.io/home-assistant/home-assistant:stable"
volumes:
- /PATH_TO_YOUR_CONFIG:/config
- /etc/localtime:/etc/localtime:ro
restart: unless-stopped
privileged: true
network_mode: host
I asked how you installed docker
ping: bad address 'google.com'
``` DNS is fucked
via this command a year or so ago
curl -sSL https://get.docker.com | sh
then surely it would fail outside of docker? Whereas it works fine
not if the host has google cached
REceieved an upgrade notification for homeassistant os today, Installed the new version, now all hacs lovelace integrations are not working. any hint? known bug?
There are no lovelace integrations
There's cards for the thing that was once called Lovelace
There's custom integrations
However, updating the OS shouldn't impact either ... check your log file
Same. All is borked here now. 😛
Some work, some dont, some timeout. Hue doesn't recover. Also takes a while to fully boot.
Haven't tried full restart with power off yet.
Does docker just re-use the host DNS settings?
Did a full restart already, no custom cards are working, even checked if the cards are still installed. any hint what log to show?
If it's really cards then #frontend-archived can help
Logs accesible from the web interface have nothing in there regarding frontend.
Mushroom is working, custom:mini-graph-card and custom:energy-overview-card are not shown on any dashboard
I don't know what it is but it feels very network related. Integrations go off and online. Devices cannot be reached.
There's a #frontend-archived channel... that can help with cards
turns out it was a broken symbolic link from /etc/resolv.conf
host doesn't use resolv.conf, which is why it was working fine, whereas it appears docker uses it.
I have no idea whats running on your machine
it's prolly lonely, you should give it a rub every once in a while to make it feel appreciated
@quick wagon
the first one says " NGINX is used as a reverse proxy, ensuring that traffic that comes in over a particular port is carried over a secure SSL connection"
isn't that what let's encrypt already does?
No, LetsEncrypt provides you with an SSL certificate. You have to use that certificate - either directly in Home Assistant or in something like NGINX
Do you want to access HA locally using the IP or homeassistant.local? Or are you happy to always use your DuckDNS/whatever domain name?
I don't think it will impose a difference, I already use my stream deck to open HA website, so I don't write the url everytime
Right, but in one case you're accessing it using your public domain name every time, which requires working Internet
In the other you're doing what you're doing today - going directly to the HA host
(you really need to learn about networking)
I really do
sorry for the ping
If you're happy to rely on your Internet connection then just use the DuckDNS add-on and do it all in HA
If you'd rather not then use a reverse proxy, like NGINX
so NGINX will allow me to use both the public and the privet domain?
I guess that will be better
I tried to set up duckdns with nginx a few weeks ago and faced a lot of problems, But I think the main issue was because of my port forwarding
because I'm using an eero mesh network connected to router/modem combo isp (which I can't bridge)
Well, at least using the NGINX route you won't break your local access
I could give you a summary of the problem if you can help
Port forward your chosen port from the router to the Eero's WAN IP, and then from the EEro to HA
yes someone told me to port 443 to 443 using eero ip and another port 8123 to 8123 for eero in my isp
then 8123 to 443 in eero app
is that what you mean?
Well, no - pick one port and forward it all the way
can you give an example please?
isp internal 8123 -> external 8123 (to eero's ip)
eero internal 8123 -> external 8123 (to HA's ip)
?
Yes
okay I'll try that, thank you
When port forwarding in my isp, do I leave the "External Source IP Address" empty?
every time I tried to update HA I got an error,
I rebooted the server and got this
What hardware are you running this on?
an old laptop, with HAOS installed directly on the internal drive
Just trying to make sure it's not a pi with sd card or SSD without appropriate power causing corruption because that looks like corrupted/dying storage
rip
runs after it
Why?
an old laptop
Because ☝️
All electronics die. Storage (and power supplies) tend to die sooner than other stuff.
so do I just get an ssd?
SD cards die very fast, HDDs die at what seems like random, SSDs usually last longer
but not all electronics truly live
MTBF is low, but never 0
MBTF is also rarely helpful to consumers ... usually
my experience is they either die in their first 3 years or they live forever
there is no in between, no.
and that means that I lost all HA automations and have to start from the zero right ?
that's why I should've backed up to the cloud -_-
Well, it's not a backup if you leave it on the computer you're backing up
Cloud, some other computer, even an external drive...
It's very much possible that only certain things are corrupted and your precious data is "fine"
Drive recovery software is worth a shot, though obviously it needs to know Linux file systems
Stop the laptop, do not write any more to the disk and plug it into another computer to read/copy its data off of it
then?
then go through it bit by bit
Hi everyone, quick question. Would it work if I installed HA to an external NVME in an enclosure (USB) on my old laptop, then simply unplugged it and ran it from a mini PC when I get around to getting a cheap one? Or would I have to do a backup/restore on the new one instead of plug and play?
Hii,
On this page https://www.home-assistant.io/installation/macos/, it's x86_64 image ? How get image for ARM (M1 M2 Chip) ??
Don't believe there is one
Okay thanks
I found solution https://github.com/home-assistant/operating-system/releases
aarch img on this link
Seems like my server didnt like update from 2022.11 to latest. I see there is a mariadb update from 10.5 to 10.11 Now there is a problem with the mariasql complaining it cant find a .pid file. I have searched online and see it is something related to innodb.
Is there a easy way to revert it to 2022.11 image?
From .err under mariadb if anyone has an idea:
023-05-30 19:46:44 0 [Note] Starting MariaDB 10.11.3-MariaDB-log source revision 0bb31039f54bd6a0dc8f0fc7d40e6b58a51998b0 as process
2023-05-30 19:46:44 0 [ERROR] InnoDB: Upgrade after a crash is not supported. The redo log was created with MariaDB 10.5.13. You must start up and shut down MariaDB 10.7 or earlier.
2023-05-30 19:46:44 0 [ERROR] InnoDB: Plugin initialization aborted with error Generic error
2023-05-30 19:46:44 0 [Note] InnoDB: Starting shutdown...
2023-05-30 19:46:44 0 [ERROR] Plugin 'InnoDB' init function returned error.
2023-05-30 19:46:44 0 [ERROR] Plugin 'InnoDB' registration as a STORAGE ENGINE failed.
trying to install it on a dell poweredge 1950 yes i have found it
You write the image to the boot disk, or install Proxmox and install HAOS in a VM... which would be the better idea on a server like that
Hi,
How to restore backup on x86_64 to aarch64 ?
Just restore the backup
It's doesn't work
i do not know how
How to what?
read the guide?
restore the backup
I wasn't responding to you there...
ok
I've the logo HA with a spinner since 10 min 😦
It can take a while on large backups
any one have any ideas on what else i can run it on
it can take a while on small backups
run what on?
there is not indicator or pourcentage bar it's normal ?
No, there's not
hoas
no, its normal
Okay i waiting
hardware like a chromebook
ask in #hardware-archived
Install Proxmox on your server
adding this to the error messages... mariadb | Caught SIGTERM signal!
mariadb | cat: /var/run/mysqld/mysqld.pid: No such file or directory
any inputs and thoughts appreciated
you talk to me ?
I see on reddit there wait 2-3 hours for 800MB backup. My backup size is 124MB and restoration run since 20 min now i'm waiting.
Nothing work, i tried to restore after config HomeAssistant but nothing happens. I need help
hi! I'm going to replace my old install (raspberry pi 4) with an i3 intel nuc. For those of use who have tried both, would you install HA as a container (Docker, LXC) or HAOS virtualized? I wonder if giving up on HA addons would feel like a chore, having to setup all containers myself, or its not that big of a deal
Only you can answer that but it probably will. Nowadays I'd always virtualize. I'd suggest you run HAOS and install additional things in a separate VM or container when needed. Personally I feel too restricted by HAOS and use a docker install but many people love it
Thanks for the no-help 🙂 fixed it myself by renaming the mariadb folder and rebuild it
I run it in docker on a i7 nuc and deploy everyting else I need myself as Docker Containers. For me it feels like i"m able to rollback certain things easier and it makes getting rid of stuff and backing it up more easy to me. As I already had a Workflow for that cause I started with some other non IOT related stuff years ago. Like Document management in Paperless which I also dockerized.
I want to buy an ssd for HA server, and found
15.5$ for 128GB
18$ for 256Gb
25$ for 512Gb
is it worth getting the 512 over the 256?
always
does HA take that much space?
also that
or adding media.. you never know
the ssd will only be HAOS
yeah that makes more sense
Like a Camera integration that takes snapshots and send them to you, you forget about deleting it afterwards an the Disk fills up
10$ 0_0
well personally i'd use a samsong 870 evo or something instead of w/e ssd is $25 for 500 gigs
but you can get away with a cheaper kingston a400 etc if you want to save $ i guess
for example, and Storage is so cheap nowadays that I would always go with the biggest viable option
a lot of them are cheap as dirt now Smart
it's Silicon Power Ace A55
eww
get a reputable brand for HA
if i havent heard of the brand i dont buy it especially for drives
my 256gb SSD cost me like $45 here in The Bahamas, retail
Samsung, Sandisk and Kingstone are not that expensive
it's a kingston
And well proven to work reliable
are you in the us?
and for The Bahamas to have SSD for $45, must be like $3 somewhere else
KSA
samsung 500gb would be 50$
I would too in a perfect world
i'd go 120gb samsung over 500gb noname
almost double the 28
still $50 is not bad
I just offloaded everything to my NAS 😄
i keep my server separate from my network separate from my smart home separate from my heat/ac separate from my cameras
like I said I paid $45 for a 256 and that's pretty damn good for here
the 512 is like $80 I think
see my price difference made me go with 256, but for $18/$25.... that's a no brainer
I found 480Gb Kingstone for same 28$
Yes but 😄 I needed backups anyways and the easiest way was a read only mounted boot drive and them mount a shared volume of my NAS to my NUC
the NAS was already doing backups to a remote location where I run a second one with almost the Same specs that just runns minio
But in general I also prefer to keep it seperate
what you guys think?
sounds fair, but even a good drive is not a replacement for an offsite backup
i have a few kingston a400 they arent the fastest but they've been reliable
they arent QLC
If you want to run your house off it without pissing off your Wife, make sure it's stable and quickly up and running again if something fails 😄
tell me about it, I'm buying an ssd cuz my hdd died on me
well at least I think it died.. I didn't check or try to fix
Well, I just bought everything twice, set it up the same way and then threw one into storage, the only difference is the backup hardware runns it's BaseOS off a HDD cause an SSD might lose data if it is not powered for a long time(talking years of storage) and I also made myself an PSU adapter for the NUC(they run fine off 12V) so I could even power it out of the DC/DC converter thats connected to my PV Battery
you do not need a NAS for it, there are cloud services that do not charge a lot for small amounts of storage, as long as you backup your integration configs as well as the main config.yaml you should be back up and running quickly
If you have a old mainboard, cpu and ram flyingaround.. and the mainboard supports raid1... and you find your a couple of hdds... you can easily install xpenology as a docker image from core.
Plug a big storage medium into your existing server and you can technically have a NAS
ye or just use FreeNAS
I do have maybe total of 4-5 laptops laying around unused
but I don't think I need a nas currently
yeah. freenas is legit . I tried to use xpenology but the update sequence from version to version was not flawless.
I have learned today:
- create mariadb backups
- create gui backups
- create container backups
- create enviroment backups
just to be sure.. my laptop with 2.5 inch hdd would fit a 2.5 ssd, both have the same thickness right?
ye
great
as long as it has a sata port
you should still setup automated backups to google drive or something
I should, that's why I asked how large are the back ups?
download kingston ssd manager and run firmware update on the ssd to ensure it got the newest optimizations.
gdrive is a great solution and the free 5gb? are more than enough for your configs
backup size depends on size of db and addons
mine are 500mb nowadays
I'm at 700mb for the Homeassistant docker container atm
I could use the free plan of google drive 0_0
just linking up the container to the gdrive then or run freebackup software or something?
and you guys have it back up every how much time?
i use a third party nextcloud backup addon in haos
And you could do like a daily or weekly backup then delete like 4 out of 4 backups per month
as soon as you reach a year you could keep on of the backups for that year and delete the rest
to free up some space
it is the updates that ruin the most. So backup before upgrades! 😛
Daily, Monthly and event triggered
event triggered?
yep
like updates?
when you unlock your front door at home 😄
bro why
Before my Scripts rum the deploy script for my containers it does backups
isn't that a lot of backing up?
I can understand daily, but after every event??
backups goes in the hidden. So it will not take much space either. With free-space at google you should have some good retentions.
not like every Hass event
doesn't that cause a delay?
weekly + before each update is plenty
or monthly if you dont change shit or care about historical data
That's what I thought
agree. i lost all historical data todayy
I have a bash script that updates my containers every week, that also triggers the backup script, waits for it to finish and send back a "ok" message
I lost all my power consumptions : (
then it continues with redeployment
sound very okay routine
good luck @quick wagon
It also sends me an email if the backup does not work but thats handled by the backup script itself
i want to enable turn on lan feature on an lg web os tv does anyone know how
Hi everyone, quick question. Would it work if I installed HA to an external NVME in an enclosure (USB) on my old laptop, then simply unplugged it and ran it from a mini PC when I get around to getting a cheap one? Or would I have to do a backup/restore on the new one instead of plug and play?
https://www.home-assistant.io/integrations/wake_on_lan/ and this blueprint https://community.home-assistant.io/t/lg-webos-smart-tv-enable-turn-on-action-for-ha-2022-2/388153
I just set this up the other day. Also your you need to enable turn on via WiFi in your TV settings.
Can someone confirm that home assistant 's docker is over 1 GB?
Yes, 1.6gb
Hello, hopefully im here in the right Channel.
I have HA running on a Rpi4 which boots from a ssd.
Now i want to Backup to an external Hdd..
How Do i Mount this into HA OS?
Is the Volumename CONFIG the way to Go?
HAOS doesn't support mounting disks like thhat
Yet?
Use a local file
I tried, it also failed
Please also copy the text of the error, and share that, rather than sharing an image of the start of the error
Error: Error invoking remote method 'enable-screensaver': No handler registered for 'enable-screensaver'
at o.invoke (node:electron/js2c/renderer_init:57:526)
Are you trying to write to the failing HDD?
if i were you i'd just try using Rufus
you can use it without installing with the portable exe
and its 1 megabyte
I'm sorry, I couldn't copy this
rusfus is not letting me choose the ssd
can you show a screenshot
it only shows usbs
click ctrl+alt+f
alt+f is nto the same as ctrl+alt+f
seems like rufus is designed to only write to usb drives
even though etcher says it failed, doesn't that look like it flashed it fine?
the error shows after finishing
those are some weird partitions lol but i don tknow how the haos is structured
I'd try it with just the .img, as in, extract it yourself, but there's other options here: https://alternativeto.net/software/etcher/
It will create multiple partitions. It might have just been stuck on one of them. Did you have Disk Management open while flashing?
it writes the partition table first, doesn't mean their contents is also finished. very much the opposite
also failed
doesn't matter if disk management open or not, both failed
sounds like disk issues
try a reboot or something
already restarted the pc
write from linux using dd 🧌
Could also be a problem with the USB adapter you're using, if you're using one
i dont think they are
it's directly plugged to pc via sata cable
No ISO
img.xz
It's a disk image
wow
Your tool may require you to uncompress it first
First, burn this CD image..
I tried both, still failed
well at that point i'd just try writing it with dd
That or booting from a Debian USB stick and install Debian just to prove the hardware
yeah you'd have to decompress it to .img
Hi all. Im having a bit of an issue. I did a restart of HA and now i cant reach it anymore. is this the right channel to get som help and input?
Yes it is
Hopefully you have either working SSH access, or a keyboard and monitor to connect to it
everything worked fine except an intergration that sometimes stop working but comes back after a restart, so i did that. Now i can see it on my router, i can connect to it and get all files throug explorer, and if i type in ip::4357 everything is fine and healty but i cant reach it via ip. And nabu casa cant see that its online
are you running etcher with Admin privileges on Windows? you could also try downloading the image file from the URL to your PC and then flash
Yes, and I tried both from url and file
both failed
in the screenshot you sent earlier it appeared some data was written to the drive. Maybe attempt to boot it and see what happens. Assisting a friend with an install recently, etcher continuously failed at 28% writing (program would crash entirely), yet, it still seemed to have written successfully (seeing similar disk results), and upon rebooting all was good to go
it fails after finishing
etcher says it failed (as it did for me), but you do see partitions being created on the drive (as did I). Might be worth a shot. If not, flashing directly from the machine (booting from a live USB) is really easy
will do
See here for a manual way: <#installation-archived message>
You shouldn't have to do that but 🤷
I think I did it
I found out it always fails after flashing and while validating
so I just skipped the validation
skip it, and ship it, haha
i have the ssh addon in HA im pretty sure...
But did you set it up to listen on a port so you can actually connect to it?
dosnt look like it 😦 and cant find the mini hdmi cable to connect to monitor....... need to buy that tomorrow then.,..... sucks, all i did was a restart
and i cant do anything from explorer right? i can see all the files
That simply means that the samba addon/container is running, I guess. You can't really do any diagnosing with that
ok.... but the SD card should be fine if i can see this : Home Assistant observer
Supervisor: Connected
Supported: Supported
Healthy: Healthy
right?
I'd have to research how it determines that but that's not a bad sign
but if i get a mini hdmi so i can connect it to a monitor and a usb keyboard, what can i do ?
Depends. Checking the logs would be a start: https://developers.home-assistant.io/docs/operating-system/debugging/#checking-the-logs
hi need some help on installing HA OS on proxmox
I enter the command: bash -c "$(wget -qLO - https://github.com/tteck/Proxmox/raw/main/vm/haos-vm.sh)" in proxmox shell
and nothing happens
what am I doing wrong here?
I just tried it on a non-Proxmox box (as I don't have a spare Proxmox host to test on) and it worked
is there some setting I need to do before running this command?
it just goes to a new line in the shell
Assuming you're following the guide step by step, just the previous steps
Try to do the steps manually. Something like this
wget https://github.com/tteck/Proxmox/raw/main/vm/haos-vm.sh
bash haos-vm.sh
If that still doesn't do anything, try bash -x haos-vm.sh
--2023-06-01 01:52:23-- https://github.com/tteck/Proxmox/raw/main/vm/haos-vm.s
Resolving github.com (github.com)... failed: Temporary failure in name resolution.
wget: unable to resolve host address ‘github.com’
root@pve:~#
Well, seems like your proxmox install is misconfigured. Show the contents of /etc/resolv.conf, /etc/network/interfaces and output of ip a
The gateway ip is usually a good choice. You can use a public one like 1.1.1.1 but miss out on local name resolution.
I can probably tell you something more specific if you get me what I asked for
That's only sensible if you're running a DNS server on that computer
Which on Proxmox is pretty damn unlikely
What about the rest?
ip a:
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: eno1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master vmbr0 state UP group default qlen 1000
link/ether 10:a0:76:db:62:1f brd ff:ff:ff:ff:ff:ff
altname enp0s31f6
3: vmbr0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
link/ether 10:a0:76:db:62:1f brd ff:ff:ff:ff:ff:ff
inet 10.10.1.123/24 scope global vmbr0
valid_lft forever preferred_lft forever
inet6 fe80::42b0:76ff:fedb:621f/64 scope link
valid_lft forever preferred_lft forever```
I dont have a /etc/network/interfaces
Try 10.10.1.1. How did you install proxmox?
from a usb stick
where do I set this?
/etc/resolv.conf
so I should change 127.0.0.1 to?
You can actually change this in the web interface as well under yournode > System > DNS
The gateway ip is usually a good choice
Try
10.10.1.1
It's almost like you were told these things already 😛
thanks
sorry... this is way above my IT grade
now it works! gosh.....
thank you so much!
Okay, I am struggling to install HAOS on my dedicated miniPC.
I have this: https://www.amazon.com/gp/product/B09B1XGNKV/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1
And this SSD: https://www.amazon.com/gp/product/B003J5JB12/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1
And a bunch of USBs. I'm trying to use Balena Etcher to write to either a USB or the SSD, but the mini PC just boots Ubuntu every time. And Ubuntu works so slowly it is basically useless on this machine (I may need hardwire mouse/keyboard). I tried to use my regular windows PC to use Balena etcher and have been struggling with that and either it fails or it just won't boot HAOS.
Not sure of what else to do here.
Put it on the storage containing ubuntu or go into bios and change the boot medium. Make sure to enable UEFI booting
So on my computer I see a section called (boot/efi) where I could potentially put HAOS onto. Will that work?
(When I open Balena Etcher and try to flash)
Okay, that's what I thought
Hmm, I have some partitions created when I installed Ubuntu onto this machine
When you write the image it over-writes anything that's there already, erasing it all
Yeah, but in balena etcher those partitions are locked
I have this SSD but it fails every time I try to flash to it
what about a live usb and booting from that?
Yeah, I'm flashing to a USB right now one more time and seeing if I can't get it to work
I think there's some weird stuff where it doesn't work if you flash from a windows machine or something
I flashed from a windows machine just fine many many times
but I've also used live Ubuntu to flash
when windows was behaving like windows
I got it to work by reflashing the USB from Linux 😐
But I don't really want to use a live USB because this one is weirdly finicky.
Gotta figure out how to get this booted from my SSD eventually
you only need it for the flashing tho
Doesn't it boot from the USB when you flash to a USB?
no
when you flash the img to a hard drive from the usb, you take the usb out and boot the image on the drive
ho
it is HA, not ho
Hi. After an update yesterday (done through the Android app) my HomeAssistant will not boot. HA runs as a VM. The boot fails when trying to find a partition labelled hassos-data - lsblk tells me there is no partition labelled that, only partitions labelled hassos-boot, hassos-overlay & hassos-data-old. How can I recover my system?
Hi everyone, I have Home assistant OS installed on a NUC Mini Pc.
I also wanted a plex server installed so I found an addon but I figured it’s not the best way to do it as I can’t add libraries from external hard disk.
What would be your advice to have HA & Plex server hosted on a NUC ?
Should I install Windows with Plex server + a VM with HA on it ?
I want something easy and not complex
I’d use proxmox
Ok I never heard of that, is it free ?
I also had installed transmission torrent client on my HA, is it possible to do the same with proxmox ?
You can do pretty much the same thing with a HAOS VM that you can do with a bare-metal install of it
Hey guys, I'm having a proxmox issue that hopefully someone can help with - sorry if this is a rookie question but I'm new to this and haven't even installed Linux before so I have no idea what I'm doing
I'm trying to follow this guide https://www.derekseaman.com/2023/04/installing-home-assistant-on-proxmox-quick-start-guide.html
And for the ip section it says if you're plugged into your router it'll grab an address and populate the rest of the info
Now that it's installed its telling me to go to a website to configure the server, but the site it's telling me to go to is 192.168.100.x, but everything else on the network seems to be 192.168.1.x, and I don't know how to access the proxmox server?
Someone had to give it the 192.168.100.x. You can open the VM's VNC console, type login and then ip a to get the address of the VM
It was auto configured to 100.x
When I had it plugged into my router, that was the default option
I did ip a and it's come back with 3 things - first is 127.0.0.1, second is... Something that doesn't make sense? And the third thing is the same 192.168.100.x ip from before
Looks like your router's DHCP server gives out 192.168.100.x adresseses. Is everything else configured with static ip adresses or do you perhaps use VLANs?
Kinda strange that your PC would get 192.168.1.x assigned while the VM gets 192.168.100.x on a "normal" network
I'm not sure what that means sorry - I'm very new to this
I do have static ip set for a lot of things, but not everything
Like my laptop and HA instance, and some Shelly devices etc are all static ip
But I don't know what vlans are?
I honestly didn't know it was possible for my router to give it anything other than 168.1.x honestly
Also if it helps my router is based in 168.1.x
Careful what you type
Oh you just leave the 192 away thats ok i was confused
There‘s no reason to use static ips on those clients
Seems like your routers uses 192.168.100.x and you statically assign 192.168.1.x.
You can statically assign the same to your HAOS if you want with ha network update ....
Using different ranges is probably not a good idea for network beginners though
I thought since my router was based at 192.168.1.x that that's what it would be assigning to everything though?
I am so confused
Thats probably your issue, your proxmox doing some subnet mask fuckery to talk to some clients in the other net
This is so confusing
Is there a way to fix it?
Force proxmox into 168.1 so I can configure the network like in the guide?
Have everything use your perfectly fine dhcp advertisements
No static stuff and you should be fine
Proxmox really likes to use/have a static ip though. A DHCP lease works if configured manually but the network editor in the UI will not understand it and mess the config up
I was under the impression ha does too
„Static dhcp leases“ are not the same as static ip settings on clients
And the Shelly's I just did so I could tell which one was which if I ever needed to change the settings for one of them
You can use static DHCP leases
That's cleaner and harder to break your network with
Uhh, how do I do that?
You configure that in your router's web interface
By not setting static ips in clients 😛
But also for now, how do I get proxmox to work? I can't access it from any of my computers
So if I disable the static ip on my laptop I'll be able to go 168.100?
How can you access proxmox at the moment?
I can't - it's just installed and all I'm seeing is a screen telling me to go to an 168.100 ip address to configure the server
That screen being a monitor plugged into the proxmox machine?
Yeah
set your laptop to a 192.168.100.x address, access proxmox again and change proxmox to whatever your lan actually is. then change the laptop to dhcp
Okay. Log in, run nano /etc/network/interfaces, change the ip there, CTRL+X, then ifreload -av. ip a should now show it.
You might also have to modify /etc/hosts the same way and also check /etc/resolv.conf for the correct name server(s).
Make sure no other device uses that ip
The x part of 168.1.x can be anything from 0 to 255 technically, right?
.1 would be the router and .255 is usually broadcast so don't use those 😄
yeah, just not the same as your laptop
Also I take it I have to change my gateway to 168.1.x too?
Yeah, easy
Good point!
gateway is usually not needed in this emergency scenario as you access the device directly
So should I keep it as 100?
your lan is 192.168.1.x right?
Yeah
then make it 192.168.1.253 for example
The gateway?
who talked about the gateway?
The gateway has 192.168.100.x and I was just wondering if I should keep it as 100 or change it to 1 to be safe?
ah
yes, change that also to 192.168.1.1
I've seen ISP modems occupy .254 or even .138 but yes in most cases its .1
kvm (proxmox) indeed wants to have some special treatment with its own static ip but just stay away from setting them on regular lan clients. use your routers config page to set entries which get handed out statically
Don't think it relates to kvm it's just that the network editor doesn't care about your dhcp entries and happily removes it from the config if you use it.
It also wants to be able to resolve the ip to the hostname or the reverse and a static ip that is also in /etc/hosts helps with that
eh, I've seen it struggle on regular debian in the same way
Ok, new issue - in the guide there's a post installation script but it's not running the bash command
A quick google says it might be because my dns is pointing to 127.0.0.1 for some reason - should I change this to 8.8.8.8 and 8.8.4.4?
you can
Or are there better options?
Your router is most likely able to resolve local hostnames for you so it's a good idea to use 192.168.100.1 or whatever the router's ip was.
For the dns?
Yep
Cool!
so now that i've got proxmox working through the web interface, i don't need the screen and whatnot plugged in anymore, do i?
Nope. You usually configure proxmox through SSH and the webinterface. The screen is useful for emergencies though
perfect - it's my main screen so would be nice to have it back now that the hard part is done in theory 😛
so when i move my current HA instance to the new proxmox machine, will it keep all the settings like the static ip and everything that i've configured?
you should make a static DHCP assignment for it in your router as well
for your HA box
Depends how it was set up before but I'd make a snapshot/backup and try it out
sweet, migration to proxmox is a success!
I have 8GB allocated for HA (in a VM) is it overkill? This is the output of free:
total used free shared buff/cache available
Mem: 7.8G 1.8G 3.9G 4.9M 2.1G 6.1G
Swap: 2.6G 0 2.6G
A little bit but it doesn't hurt. Just do ballooning (if possible) and it'll be fine
I want to keep as much ram free for my host OS. Thanks.
I wouldn't go lower than 3g in your case as it looks like 2g~ are used
I was thinking to put it to 4. I decided to leave it for now unless I actually need it for something else.
Doesn't matter
hmmm
are you contemplating your existence because you asked the same question earlier?
nah, just the response lol
From a personal perspective, I prefer to update the operating system before updating Core.
hey guys, i moved my HA install yesterday from my windows machine to a new proxmox setup and made a backup and moved everything over, but i just realised most of my addons didn't copy over
i'm trying to reinstall them, but i've installed them, restarted HA and am trying to start them but they're not starting - there's just a red circle in the top right that when hovered over says "this addon is stopped", and there's none of the usual cpu usage or anything to go with it
with mosquitto for example, i try to start and nothing happens, and i check the logs and there's nothing there even saying it tried to start and failed - the logs only go until when HA started back up
does anyone have any ideas on how to fix this and get my addons going again?
did you check the addon logs?
@quasi wedge I converted your message into a file since it's above 15 lines :+1:
Restore from the backup again it sounds like maybe something went wrong
It should be a 1:1 copy once you restore from backup
bah, i just got it all working last night 😦
is there any way to not do a full restore?
you can clone a VM in proxmox and play with another version with one offline if you want
or you have to get manual
i can also do a partial restore from a full backup!
and i can specially restore the addons
yeah like I said you'd have to get manual
bah, i have bad luck with tech so have a feeling if i were to do a partial restore, i'll only find something else down the line that's not working that'll need a full restore anyway
i'll just do the full restore
every time I change something I name the backup before I make the change
how do i know when the restore is done
i hit restore and all that happened is i lost connection to HA, but there's no progress or anything to see how it's going?
do i just wait and keep trying homeassistant.local:8123 until it starts working?
ah, yes - eventually homeassistant.local:8123 loaded and it looks like the addons are all there - will test it out and see where things are at
remember you need a full restart of the host VM after a successful restore from backup
ah, good point - will restart from proxmox now
should do the same thing in haos when you choose reboot not restart
I used the homeassistant docker container to install HA on my NAS. Works great, except I apparently can't use add-ons this way (???), and I need to install Mosquitto Broker. It's instructions only provide install instructions via add-ons, so I'm a bit stuck at the moment (obviously I'm a HA noob). Halp?
so install the container for it?
got it. So the answer is just run it separately. Can do.
that's all an addon is really
first I should ask, am I correct in assuming that I cannot get Tasmota integration added without MQTT (which requires a broker)?
I have some Tasmota devices that I'd like to manage via HA.
or is that a better questions for #add-ons-archived ?
yes im pretty sure you have to have MQTT working to have tasmoadmin
okiedoke. thanks! off to get the broker working.
got the broker installed which allowed me to get Tasmota integration added... I assume then that unless the individual tasmota devices are set up to talk to the broker, that HA won't detect them?
n/m yes that seems to be true. Got it working.
how long does it usually take to restore a full backup?
I'm setting up my new mini pc (using proxmox), the VM has 50gb of storage, 4 cores, 8gb ram
I started restoring last night and went to sleep. Still hasn't finished 8 hours later
The backup is only 2.5gb
So finally it worked, now I'm having trouble figuring out how to setup the new proxmox vm to have the same IP address the HA had on virtualbox
Any suggestions?
Update the static DHCP lease on your router?
Yes indeed haha, been a long day
I also changed it in HA itself under System>Network
Hi Quick question if you don't mind.
I run HAOS on a virtualbox vm. I have 2 external USB sticks for zigbee and z-wave respectively, which are presented from debian through to the vm
I plan to move the VM to new, more reliable hardware - still as a debian-based virtualbox VM.
What mess am I expecting to make of my connections to my external zigbee and z-wave networks? Should I just be able to represent them to the VM and expect everything to work? Or am I looking at a nightmare scenario where nothing works and I have to start from scratch?
Thanks
None, if you were sensible and used the /dev/serial/by-id/ paths thhen everything will just work
Context:
Hey. I previously had HA setup as a dock on raspberry pi and functioning as intended. While securing I made more limitations for access while attempting to improve safety of my system. It did work all good for a while, but after that my internet setup was changed (modem removed) and I lost any means to connect to it anymore. I let it be for few months, but now I wanted to gain access again. After modifying files to remove that blocking, I managed to login into system. For some reason supervisor wasn't enabled while observer reported that everything was okay. This prevent me from using any add-ons, updating to newer version etc. either by UI or cli commands. All attempts to repair, check and update just failed.
I'm not sure what was wrong, but I got fe. "/supervisor/options return code 403" errors etc. into log. I took backup and decided to reinstall latest version and intended to restore my backup. I downloaded new ocker version, set it up on docker and launched it off. First configuration window didn't offer me a option to restoe backup as seen on on some guides, so I setup some temporary user.
Used installation method https://www.home-assistant.io/installation/raspberrypi#install-home-assistant-container - update tab
Problem:
I still can't access supervisor or addons. Attempting to enable supervisor from integrations guides to modify configuration.yaml which would require add-ons (that option doesn't exist anymore in settings) and observer site claims "Supported: Unsupported".
`Logger: aiohttp.server
Source: /usr/local/lib/python3.10/site-packages/aiohttp/web_protocol.py:403
First occurred: 1:10:02 PM (7155 occurrences)
Last logged: 3:18:23 PM
Error handling request
Traceback (most recent call last):
File "/usr/local/lib/python3.10/site-packages/aiohttp/web_protocol.py", line 332, in data_received
messages, upgraded, tail = self._request_parser.feed_data(data)
File "aiohttp/_http_parser.pyx", line 551, in aiohttp._http_parser.HttpParser.feed_data
aiohttp.http_exceptions.BadStatusLine: 400, message="Bad status line 'Invalid method encountered'"`
If you want the Supervisor and add-ons you needed to use Home Assistant OS
That doesn't make sense. Why it would suddenly start requiring that OS?
The only other option is Supervised, which is a clusterfuck and to be avoided
I had HA version from march and I originally set it up at the end of last year
because HAOS is the only one that supports addons, otherwise you have to do it yourself
And thne it suddenly was butchered?
any ideas of any way i can get IKEA Dirigera hub working with home assistant?
That'd be #integrations-archived
thanks
I can tell that I did have it running in on container and it stopped working at some point.
HAOS uses Docker
Alright, either way. It still doesn't work and I didn't reinstall os.
So why it behaves like this?
If you want add-ons you HAVE TO USE HAOS
based on what has been said. I had HAOS since the start.
if you manually downloaded a new docker version you ain't using HAOS
and the link you said you followed and posted was to a container install
has anyone experienced getting error "400: bad request" when trying to proxy through nginx?
directip:8123 works fine, even domain:8123 works fine (port forwarded), but trying to use nginx so I dont have to enter the port and make ha.domain proxy to directip:8123 and I get 400: bad request
note, Im using the same nginx configuration for other apps on other ports and it works fine
thanks so much ❤️
So I'm still working through my overall setup and considering moving to the Docker installation, rather than Core.
I found this video that helped explain a great deal about how to use with the Docker installation and still access and use things that are otherwise in the HACS (and more, actually): https://youtu.be/DV_OD4OPKno.
However, while I have Docker installed on my linux HA dedicated laptop now, I'm not very fluid in how it works yet.
Does anyone know if I can install Docker Containers for things like Node Red for example and still make them work with HA Core?
Or if I'm going to install the NodeRed Container, do I have to install HA in a docker to make them work together?
I'm more interested in the general question of whether I need to migrate from Core to Docker, not specifics of NodeRed in particular... Anyone know how that works?
Does anyone know if I can install Docker Containers for things like Node Red for example and still make them work with HA Core?
Yes, you can
Migrating from a venv (pure Core) install to Docker is normally pretty trivial. You'll have to update any paths that refer to the current on disk location of files/folders to how they'll appear inside the container, and then you can simply start it with your existing config folder (https://blog.ceard.tech/2020/10/ha-venv-to-docker)
If you currently do SSL directly in HA then you'll have to map a volume for the SSL certificates, or save yourself pain and use Traefik as a proxy since it can handle SSL for you
I'd 100% recommend Traefik
(that's only partly because I use it myself)
lol. Got it. A the moment, I'm not accessing it remotely - onyl from within my own network, so I haven't moved up to SSL yet. I'm likely to going to put up PfSense before I do that - so I thought I could handle that when I do the pfsense part -but I will look into Traefik in case I have that wrong.
You can handle it then - Traefik is awesome since it's tightly integrated with Docker
And I do have it running in a .venv at the moment - so looks pretty straight forward to move it up
Oh, and use Docker Compose - don't use docker run
A compose file gives you YAML you can back up, and makes it much easier to modify containers
I didn't even know about Docker Run yet... I only know about compose 😉
Good... you'll save yourself pain
I'll also suggest you look at Dozzle for an easy way of seeing your container logs
If you set up healthchecks for your containers then Container Mon is good for alerting you about unhealthy containers
how can I install HA on a hikvision camera server?
Does that run Linux and support running Docker images?
(or support running VMs)
tbh idk since it's not for me, I'm asking on behalf a friend who has that server and want to start using HA
but i'll see
Well... you need to find out these kind of things so we can help you repeat what we tell you 😛
this has all the makings of a disaster
so atleast it should be linux and not their propraitory OS
To run HA the device needs to be (ideally Intel/AMD) 64 bit and either support virtual machines or run Linux and Docker
The odds for yourr friend are not good
okay then, thanks a lot that’s what I needed to know
Does the hass container really need to run privileged?
Well, yes, but no, but officially yes
Without it you have to pass devices yourself, and get it right. Some people struggle with that
same story with host networking?
It's for *cast things and the like afaik
and discovery
Not using host networking breaks any integration using mDNS/SSDP/broadcasts/etc
I assume, unless whatever it's trying to talk to is on the same Docker network... I wonder if it's possible to put a host interface on that network... yes, I'm learning Docker on the fly here
Sure, but mDNS etc are used for talking to other things on your network, like Chromecast/Google Home devices, or Sonos, or ...
Thus the thought of putting a host interface on the network
Actually it sounds like macvlan was tailor-made for this purpose
@languid bough when installing haos vm via your script is there anything special that needs to be done for reliable usb passthru? (for a zigbee stick for example)
sorry if this is wrong room on the server.. I am trying to install home kit infused dashboard remotely. I need to get three folders onto that network .. but obviously I am not locally connected so i cant go through the /config folder on file explorer on my comptuer... and file editor from hacs does not let me upload folders .. only one file at a time. Any suggestions ?
thank you
I was able to get all the HACS downloads done.. just been struggling on getting the Homekit infused folders onto the HA network
the ftp system seems to be a little above my knowledge base unfortunately .. kind of greek to me lol
Now I recall why I didn't really want to reinstall this. Used balena etcher and latest image (haos_rpi2-10.1.img.xz). Added 8.8.8.8 and 8.8.4.4 as DNS options to get past first cli error. After that I left this overnight to wait on "(this can take up to 20 minutes)". Even after restart it still gives these errors. Home assistant observer doesn't show any problems, ha su repair doesn't make a difference, and ha core rebuild command gives error.
INFO (MainThread) [supervisor.jobs] 'ResolutionFixup.run_autofix' blocked from execution, system is not running - CoreState.SETUP
ERROR (SyncWorker_2) [supervisor.docker.interface] Can't install ghcr.io/home-assistant/raspberrypi2-homeassistant:2023.5.4: 500 Server Error for http+docker://localhost/v1.42/images/create?tag=2023.5.4&fromImage=ghcr.io%2Fhome-assistant%2Fraspberrypi2-homeassistant&platform=linux%2Farm%2Fv6: Internal Server Error ("Get "https://ghcr.io/v2/": context deadline exceeded (Client.Timeout exceeded while awaiting headers)")
WARNING (MainThread) [supervisor.homeassistant.core] Error on Home Assistant installation. Retry in 30sec
INFO (SyncWorker_0) [supervisor.docker.interface] Updating image ghcr.io/home-assistant/raspberrypi2-homeassistant:landingpage to ghcr.io/home-assistant/raspberrypi2-homeassistant:2023.5.4
23-06-03 09:58:08 INFO (SyncWorker_0)
INFO (SyncWorker_0) [supervisor.docker.interface] Downloading docker image ghcr.io/home-assistant/raspberrypi2-homeassistant with tag 2023.5.4.
INFO (MainThread) [supervisor.jobs] 'MountManager.reload' blocked from execution, mounting not supported on system
More extensive list of errors https://pastebin.com/DZ4T8bbP
those are DNS errors
that's unfortunately not HA
also you're running on the slowest hardware supported
Dns/gateway issues yes
Hi. After an update a few days ago (done through the Android app) my HomeAssistant will not boot. HA runs as a VM. The boot fails when trying to find a partition labelled hassos-data. lsblk tells me there is no partition labelled hassos-data, only partitions labelled hassos-boot, hassos-overlay & hassos-data-old. How can I recover my system?
reimage, restore backup?
Is there documentation for getting the backup off, I assume, the hassos-data-old partition and imported into a new install?
Has anyone tried to install and run the home assistant server on an ipad? If it run run on a raspberry pi, I guess and ipad from 2020/21 will also have sufficient hardware required to run the same. I have such an ipad lying around and would like to use it for the same purpose. I can keep the ipad plugged in for continued battery charging and running the server. Does anyone have any thoughts on this?
good luck
What’s a computer - apple
have fun porting 😂
you don't even need to port. fun
I think my sd card I had my HA on got corrupted. Will the raspi imager software fix it if I flash the OS again ?
nope
once it's corrupted you can reflash but most likely the microSD card is failing at this point
even if it does work you should take backups and be weary about trusting it from failing again soon
iPads don't so easily run custom software. If you have a Mac, you can try to compile the source in Xcode, and even if you manage to do that, you'd have to recompile and recopy every 7 days unless you have an active Apple Development account.
an exciting prospect, running HA on an ipad over wifi 🙂
Lol, homeassistant.ipk
Yeah, obviously not a good choice.
that ain't even remotely interesting to me LOL
Flashed it again and it seems to be fine now, maybe it wasn’t badly corrupted or something or maybe it wasn’t corrupted at all idek
Any software I could use to check the health of it does anybody know ?
Even if you port it, I can’t imagine iOS playing nicely with its sandboxing, it would probably have to be jailbroken to have a chance at all really