#installation-archived
74689 messages · Page 86 of 75
like for example, my homeassistent docker dir has "data docker-compose.yml "
and I use volumes: - ./data:/config
Any network experts on at the moment?
@valid heart ok lets get a few things
did you set up TLS? if not you should be using HTTP and not HTTPS
you may have to clear web browser cache / data if it keeps redirecting you
Not on reflashed install and don't think I did originally. I do use NabuCasa but don't know if that is relevant
lets ignore that and use local ip
http://192.168.1.5:8123 or whatever your HA is from a lan scan
if its not showing up after a fresh image, you have some issues
@raw hedge
docker run -d \
--name Minecraft \
--restart=unless-stopped \
--privileged \
-p 25565:25565
-e EULA=TRUE \
-v /volume1/docker/Minecraft:/data \
--network=host \
itzg/minecraft-server:stable
this is what i came up with seems okay to me
@undone sentinel To format your text as code, enter three backticks on the first line, press Enter for a new line, paste your code, press Enter again for another new line, and lastly three more backticks. Here's an example
Don't forget you can edit your post rather than repeatedly posting the same thing.
For over 15 lines you must use a code share site such as https://dpaste.org/ (pick YAML for the language), https://www.codepile.net/ (pick YAML for the language), or https://paste.debian.net/ (pick YAML for the language).
You almost certainly don't want
--privileged \
@valid heart https://www.iwaxx.com/lanscan/ may be of use to a nice gui lan scanner
okay removed
next step would be to change dns servers on the router
like try cloudflare's 1.1.1.1 or google's 8.8.8.8
Downloading now. StaticIP is 192.168.68.121/24. Can connect to http://192.168.68.121:8123 and get to HA. Chrome says its unsecure
Chrome says its unsecure
That suggests you're usinghttps://nothttp://
Yes, can sign in
ok so you have access to the HA gui now?
Yes
ok so its working now?
No. Supervisor can't update and there is a warning under System saying Your installation is unhealthy. Log has a couple of errors and a warning. Warning - 'Updater.fetch_data' blocked from execution, no supervisor internet connection
ok so now we can look at the network issue
do you know how to set a static ip / change dns on HA?
Shall I post log errors? Sorry, there a android/windows version of lanscan
lanscan was only so we can get to the HA GUI
Yes. Under IPv4 its set to static and IP address 192.168.68.121/24
clearly you have layer2 connectivity since you can access from your pc
This is also set as static on Deco Router
and the gateway?
change dns to 1.1.1.1
Done. Do I need to reboot host?
not sure with HA OS, go ahead to be safe
Ok. DNS set and rebooted & signed in again. Cache cleared as well
Similar warnings in the log
docker run -d \
--name Minecraft \
--restart=unless-stopped \
-p 25565:25565
-e EULA=TRUE \
-e TYPE=PAPER \
-e SERVER_NAME=Tsuna's Server \
-v /volume1/docker/Minecraft:/data \
itzg/minecraft-server:latest
or is this bad form to do so ?
@undone sentinel idk, I would add them manualy
environment:
- DEBUG=1```
all e flag is for environment variable
you can ignore it, declare environment under the service name and start listing them
generate it without the -e and add manually or try it and see what happens?
I dont really use environment variables in k8s unless its for options in the container, I patch instead with kustomize and a patch file
even then its for placeholders 😛
oh how I have not used the cli in a long time and i just use definition files
Sorry to be dumb, created pastebin, how do I link to it?
Save it, share the link
is... your date and time correct?
It looks an hour out of date. I'm in UK. Currently 17:00
I feel like your system time is incorrect
do a fresh reinstall
do not restore from backup
do not pass go and see if its fine
Ok. I never managed to restore a backup as it says no internet connection.
without static ip
This is what I followed last time. https://www.home-assistant.io/common-tasks/os/#flashing-an-odroid-n2
Is that correct steps to follow?
Ok. And presumably take out of the Deco router so its automatic?
This is basically a fresh install bar the static IP as I've not been able to do anything else
go back to good old dhcp and see if it works there
Ok. Will follow the steps and revert back once done. Thanks for the help so far
and if this fails, try a older image and or the new beta ones
you are having time / ssl related issues
Is the Blue a N2 or N2+?
Hey all you HA smarty 👖! I'm switching from HA OS over to Container. Just makes more sense w/ the direction I'm moving in my homelab. I'm trying to set up Home Assistant to record to a MariaDB db. Having some issues. Here's my current configs. Any help would be greatly appreciated!
@viscid cave posted a code wall, it is moved here --> https://hastebin.com/fohilomuwa
HA configuration.yaml file:
recorder:
purge_keep_days: 30
auto_purge: true
db_url: !secret db_url
HA secrets.yaml file:
db_url: "mysql://db_user:db_password@192.168.0.101:3306/ha_db?charset=utf8mb4"
192.168.0.101 is the address of my Raspberry Pi (aka. host that I'm running both the HA Docker container and MariaDB Docker container on)
Currently, in my HA logs, I am seeing: "sqlalchemy.exc.OperationalError: (MySQLdb._exceptions.OperationalError) (2002, "Can't connect to local MySQL server through socket '/run/mysqld/mysqld.sock' (2)")"
I'm guessing this is because I cannot connect to the mariadb container for some reason?
Sorry...I forgot to use a pastebin for my codeblock. Here's the most up-to-date version: https://www.toptal.com/developers/hastebin/ewukisewov.http
...added a ports: - "3306:3306" option to the mariadb container just because the docs said to.
What do you mean by this?
So I shouldn't use the host IP of my Raspberry Pi (I presume my HA container when running in host mode also uses this same IP)
So remove the ports: - "3306:3306" from my db container declaration?
First try to access the db using the service name of the database
So in secrets.yaml, try changing to:
db_url: "mysql://db_user:db_password@mariadb:3306/ha_db?charset=utf8mb4"
FYI: mariadb is my container_name
Oh...I didn't even realize those were different!! So good to know! I missed this while reading through the Docker docs.
I'll try that.
@raw hedge Is that best-practice when referencing containers from others? Like reference the service name rather than container_name?
For like IP-stuff?
Give each stack its own database
Only give network access to the DB to what needs it
Like I would make a stack, make a db network that only ha and the database software are attached to
Don’t open ports to the DB to the rest of the world
So, in my current stack, I have mosquitto (MQTT), Zigbee2MQTT, InfluxDB, MariaDB, and HA. In that Docker Compose file (Portainer stack), I have a section specified as below:
networks:
default:
name: scoobydoo
external: true
I don't expose port 3306 on the db (MariaDB) container.
Stick to making it work with db as the host name first
Full recorder integration is:
recorder:
purge_keep_days: 30
auto_purge: true
db_url: "mysql://db_user:db_password@db:3306/ha_db?charset=utf8mb4"
I'm getting this error in my HA logs: "Unknown MySQL server host 'db'"
I don't think defining db in my HA configuration.yaml file (under the recorder block) is helping
I even tried removing the port of :3306 thinking maybe that was causing an issue, still seeing the same error and it doesn't seem like the MariaDB recorder integration is working.
Do you (or anyone else) know how to make HA Container set in host networking mode work with a MariaDB container (bridged) for the recorder integration?
Now that I am not driving
Make a db network
Apply it to ha and your database in docker compose
networks:
- db
End of compose file add same thing again
I'm following what you're saying, but doesn't setting a default network (at the bottom of my compose file) allow all containers in that compose file to "talk" to one another?
Or am I misunderstanding how default networks work?
I ran docker create network scoobydoo to set up my initial network. Then at the bottom of my compose file, I have:
networks:
default:
name: scoobydoo
external: true
(I guess the lingo Docker uses is "top-level" networks key.)
would it be possible to get rid of sqlite and fully use postgres with homeassistant?
Yes
See the docs ☝️
i did check it and setup recorder but sqlite is still used.
purge_keep_days: 7
db_url: !secret postgresql_database_url```
If you're having problems with your updates to your configuration:
- Check the troubleshooting steps
- Check your log file - remembering you may need to set logger to
infoordebug - Explain what the problem you're having is - sharing configuration, errors, and logs
You need to define the network on both services
oh, i got it wrong. The recorder integration is responsible for storing details in a database, which then are handled by the history integration.
is there a way to keep configuration itself (everything that is touched by lovelace?) inside postgresql database?
No
Can anyone tell me where the lovelace configuration file is? I trashed my HA OS installation and want to save as much as possible
It's in .storage/ in your config folder
@humble mirage Thanks
My evening lights just turned on
Seems like everything is working except the frontend
Before I reinstall, is there anything I can do about this https://community.home-assistant.io/t/cant-load-the-frontend-anymore-addons-still-work/420819/2 ?
I have been unsuccessful till now
You appear to have a networking issue, hopefully somebody who knows HAOS will be able to help troubleshoot that
Hmm I just enabled a lot of ports in my router. No idea how that would break something though
"a lot of ports" 
Yeah I am currently playing around with Hyperion + diyHue + HA
Hacky stuff
Many ports on the same host
I was hoping that Hyperion could find diyHue on the HA host better like this
Port forwarding isn't going to help with anything like that
I don't know a lot about networks really
I would rather learn about the Linux kernel that about networks
Same with databases 😄
Databases are the worst
Databases are no harder than operating systems
Maybe, but 10x more boring
btw while I am here: Is there some way to control HA entities via UDP?
No
I am asking because MQTT is not really suitable for syncing light sources like with Hyperion
It would be really nice to have to UDP access point for that
You'd have to build something that could provide that
But I guess since HA talks to its addon containers via MQTT that is kind of pointless?
Haha
I just had the notion
Guess it is only for zigbee2mqtt then
Which is a problem
Or anything else that uses MQTT
Zigbee devices can't be controlled fast enough with zigbee2mqtt
So what does HA use to talk to its addons?
Well, add-ons are just software running in Docker, so whatever
I guess simple HTTP requests
If the software uses HTTP, sure
Some things do, some use MQTT, some use a custom TCP protocol
Well, no
The whole point of HA is that HA handles that
You talk to HA, HA talks to everything else
Plus I would have to write a Zigbee interface that doesn't need MQTT
What? Why?
Yeah but that is too slow... I guess it doesn't really makes sense for light syncing
You do realise that ZHA exists?
Yes, the built in #zigbee-archived integration
Ah interesting
That would probably perform better
So I would just have to find a way to expose the light group to Hyperion via UDP
The HA API uses TCP only I guess?
Yes
That is a shame
UDP for lights would makes sense in that case
For everything else not obviously
TCP is fine for that, if you don't set up a new session every time you send a command
Open a session, keep it alive
Hyperion uses UDP for controlling Yeelights for example
I think there is a reason they do
It's fine if you don't care if some messages are lost
The overhead is much lower
No for syncing I don't care
You send like 10 messages per second
Who cares if one gets lost
It must be snappy
HA also really isn't intended for that kind of use, so that could prove ... interesting
Yeah I know. That is what I am saying. I was really disappointed that it didn't work with HA.
Another issue is that some Zigbee manufactures like Ledvance only allow one pairing connection. So I couldn't pair it with HA and deconz
Well, more that HA isn't intended for near real time lighting control like that... you may find that there's going to be issues trying to get it to do that (beyond the API question)
Probably. I guess will just use the WLAN solutions of Hyperion and just not bother anymore. It works really well with Yeelights and MCP8266s
I don't have to have Zigbee for everything
But in theory it could be faster
But what would be the issue? I am curious
You mean because there is too much concurrency going on in HA?
Too many containers doing many things
Well, I've no idea what's going to happen if you try to send tens of light commands a second
Maybe it'll be fine, maybe it'll grind your Pi to a halt
And... you're not getting HA to do that
Hyperion talk to a fake diyHue bridge which is talking to HA's Zigbee devices via MQTT
All run on the same Pi4
It works, but it is laggy
Somebody once wrote a little automation to randomly flip one of ten input booleans, which would then trigger the flipping of another one. It didn't take long for that to grind to a halt
That is why you use UDP
UDP ain't magic
If the Pi is busy, it will not process the requests I guess
TCP will queue up and resend everything
Sure it isn't, but maybe at some point someone will find a solution. Maybe it won't work on the Pi4, but the Pi5 maybe
Whatever I will enjoy my laggy ambilight now haha
Thanks for the help earlier!
Am I the only one with some fast pi4s? With all of the things I am running non it?
I’m starting to think it’s because people keep using SD cards instead of external drives
SD cards are convenient, and many people forget that they're also usually not designed for a heavy write life
And they are slow
Nothing worse than someone trying to use HA for something is was never intended for and then complaining....
Hey guys. I have a raspberry PI, but no access to a microsd card reader (or adapter) is there a quick and easy install route?
Nope.... 🙂
poop
Thumb drive usb boot?
Oh I could do that, is that a possibility?
Is there a way to boot from the USb then xfer from the USb to the SD card?
Backups?
No. i mean so i dont have a USB drive hanging off of my pi 😛
Hello, I am not sure if I am in the right room or not, but I am new to HA, and I need some help setting up a device on my desktop.
@calm shoal a device?
Yes, so I have a few relay switches, that I have migrated over from Tuya, and the only button I have to make it work is Activate.
I would like to change it to an actual button that I can toggle and see the button show on/off
automations or possibly frontend?
I'm installing haos as a VM in proxmox, and I've finally gotten to being able to navigate to it locally in a browser but it seems to have just... stopped? I hit the icon for the log and it hasn't spit out any errors, it just sorta stopped
proxmox tells me it's still using a significant amount of CPU and memory, but the log has completely stopped
it got as far as:
[supervisor.host.manager] Host information reload completed
Can you enter anything on the terminal for HAOS when viewing through proxmox?
I can
Trying to set up & configure HA Container w/ MariaDB database (for recorder integration). I'm having some issues. Could be my approach? Here's my current docker-compose file:
https://www.toptal.com/developers/hastebin/roroyapawa.http
it looks like it might potentially be the DNS issue, but the ipv4 address has the correct DNS server as provided by my router. Docker doesn't, but I don't see how to change that
My hang-up is currently trying to set my homeassistant container as host network mode (so that HA can auto-discover network devices, etc.), while leaving my db service / mariadb container accessible/networked only to the homeassistant container.
Earlier today, @raw hedge suggested I define top-level networks (at the bottom of my compose file) and then reference those networks at a service level (https://docs.docker.com/compose/compose-file/compose-file-v2/#networks).
Problem is when I try starting this stack in Portainer, the Home Assistant container never starts and I get some errors.
Could it be because one cannot set network_mode: host on a container and then also specify additional networks under the service-level network key? i.e...
services:
some-service:
networks:
- some-network
- other-network
That config
You can try a restart to see if that works
If anyone has any examples or would be willing to share their stack/compose file for running HA as a Docker container in host networking mode, but also still be able to reference my db container's hostname in HA's recorder integration, that'd be awesome and super helpful!
I've restarted a few times now
Uploading an example
Thanks so much @raw hedge . I tried reviewing your earlier comments and reading through some of the Docker's docs and thought I have it set up correctly, but I'm sure I'm missing or misunderstanding something.
Only thing is that you don't have any network_mode: host entries.
I think my config follows yours, as you have yours formatted.
It should still work, or used to
What I'm trying to do is also place my homeassistant container in network host mode.
Hmm. Super odd.
does anyone know how to get the amplifi alien install on the HA
Unless it was docker swarm and overlays
I'm only using Docker standalone w/ Portainer
Let me check on it
Thanks @raw hedge . You're a rockstar!!
I don’t have this issue with k8s 😮
:p
Yah… I think it might be host mode and overlay networks
Yep
Damm.. more reminders of why moving to k3s Kubernetes was worth it
You have to expose the port and use the ip :/
I have an idea
May or may not be able to use a bridge network on a pod that is using a macvlan network @viscid cave
Probably easer for now to be lazy
I’ll have to try this tomorrow when my brain’s a little fresher. Right now that sounds kind of confusing. 😅
Kinda is for most
If you have any other epiphanies or it wouldn’t be too much to ask to draft up an example, I’d sincerely appreciate it!
I did this on a docker swarm with a pool for each node before realizing docker swarm sucks :p
@raw hedge Just found this: https://stackoverflow.com/questions/60998569/docker-compose-use-both-host-and-bridge-network
I’ll have to read through it tomorrow.
@raw hedge I'm back trying to get Blue up and running again if you or anyone else able to help.
I am trying to follow this method https://www.home-assistant.io/common-tasks/os/#flashing-an-odroid-n2
I've set bootmode to SPI and when starting, I see PeitiBoot menu as per Enabling USB drive mode section in guide
However, when I exit, there is a message
Welcome to PetitbootFailed to lauch petitboot, dropping to a sheet.
followed by
sh:can't access tty; job control turned off (which is referenced in guide as ok to ignore.)
This happened last time but I was able to proceed so ignored, but is there an issue here? Also the # prompt indents across the screen after each command is entered. The commands all seem to work but seems odd its moving. After the two messages above followed by ls dev/mmc* the # prompt is now half way across the screen
Is this an issue, or should I just proceed with install (thats what I did before and where I was at yesterday)
@valid heart did you already go for it?
No not yet. Was just looking on Odroid forum about updating PetitBoot and wondering if I should do that first
What do you think?
no idea, I just stick to pi4s or intel
Good morning! Thanks again for the help last evening! I wanted to circle back around and try understanding what you meant by maybe trying a macvlan w/ bridge network.
yahh... ignore that for now as its complex
It is?
Were you essentially suggesting to try creating a bridge network and assign that at the service-level on the homeassistant & db containers and then assign a macvlan as well to that container?
I think docker will let you attach a bridge network to a container that is attached to a macvlan network
or that might be overlay2 network
I read last evening that expose: only listens to the port at runtime and does not actually publish the port.
im trying to pair aqara roller shade driver, i have conbee usb stick and i run deconz, i cannot see it there. the device is in pairing mode
expose / pubish
tired to install zigbee2mqtt but im getting 502 : bad gateway when i try to start the web ui of it :c
sorry, my mind is thinking in k8s
Ah ok.
Here's where support was added for the Aqara roller shade driver. Maybe there's some help/ideas in the ticket: https://github.com/dresden-elektronik/deconz-rest-plugin/issues/5330
@raw hedge Sounds like I maybe have only two options then...
- sadly say goodbye to supporting HA discovery and manually configure integrations moving forward. This already works then by my initial setup by using bridge networks and publishing port
8123on thehomeassistantcontainer. - Set every container on my host to
hostnetworking mode.
Thing that sucks w/ option 2 is it seems less-secure, and even more so, I don't want to pollute my host w/ all of the published ports.
also, why im getting 502: when trying to start zigbee2mqtt ?
Have you successfully started it before? Are you running this as a Docker container? Can you provide any logs?
#zigbee-archived can help with that
Any Odroid experts out there? Looking at PetitBoot upgrade, looks like I need to use sd card but don't have reader - I was using micro-USB port to reflash HA. Not sure I can use that method on the Odroid itself? I have sd card but not reader, or have usb drives I could use.
Unless someone recommends leaving PetitBoot with error as is, I'm thinking I need to update. Do I need to buy a SD adapter or can one of the above be used?
bare in mind im a noob, never started it before, not a docker, it shows head to the configuration site of zigbee2mqtt
Head over to #zigbee-archived
oki
pollute? lol you have no idea... just being on discord, browsing the web opens many ports
Hi, I had a RFXcom 433 E and recently replaced it by a new 433 XL (that allow support for new devices not added to 433 E sadly).
My previous 433 E was declared with "device": "/dev/serial/by-id/usb-RFXCOM_RFXtrx433_SERIAL-if00-port0" (as seen in the "download the logs" from the integration menu...
How can I change the device to use to the new one "/dev/serial/by-id/usb-RFXCOM_RFXtrx433XL_NEWSERIAL-if00-port0" ?
Sadly, adding the new device conf to the yaml configuration file and restarting doesn't change my actual configuration used by homeassistant.... 😢
@valid heart I think leave it as is.
Reading around, I've seen comments that HA and Petitboot don't get on, but it isn't relevant for running HA on the N2+ anyway so looks like it doesn't matter (seems strange that is seems to work in all the instructions I've followed though). I've also asked for help on Odroid forum, as I've done the reflash succesfully (I think) in the past but it didn't fix the original supervisor internet issue I was having and won't let me install a backup.
Happy to leave it and reflash, but feels like deja vu from first time around and don't know if this is relevant. The HA issue was diagnosed as an network issue from the help yesterday and there was a discrepancy with system date/time being wrong on N2+ (can post the pastebin again if it helps).
Ok. Odroid forum confirmed just to ignore. It just drops to the shell because of a bug in Petitboot when it meets an OS installed in a storage. So I'll go ahead with the re-flash with my fingers crossed, but will come back for help before I do anything is supervisor issue remains
Good point. I guess if I'm running this box on my local network, it's secure-enough since I'm not opening every port on my router.
@raw hedge So out of those two options, which is the better move in your opinion? Option #1 or option #2?
none of those?
Haha well then what do I do?
what is stopping you from using localhost or the ip with ha in host mode and open the database port?
Use host mode networking for HA and you'll save yourself a lot of future pain
I thought I tried that this last evening when we were chatting.
If you do anything else then you're running in an unsupported configuration, and many integrations will fail to work as expected
This would be ideal! Just have been struggling to then properly connect to other containers such as mariadb for recorder integration.
Have you ever properly gotten a homeassistant container in host networking mode to properly "talk" to other containers...i.e. a db container for like the recorder integration or something?
My database lives on a remote host
Issue I was trying to debug w/ @raw hedge was that those containers like mariadb, influxdb, etc. are all in a bridge network.
I'll try what maw suggested again. Maybe I missed something last evening. So, my approach is configure homeassistant container to use host networking mode, then set my db container to use a bridged network (along with my other containers). Then in my db container, publish port 3306 for default MariaDB port. Then in my HA configuration.yaml file, set:
recorder:
purge_keep_days: 7 #or 30
auto_purge: true
db_url: "mysql://db_user:db_password@hostname_of_container/db_name?charset=utf8mb4"
Any reason to not publish (expose) the required ports to the host?
Does this sound like the right approach?
Not necessarily...just have another MariaDB instance (container) running as well on the same host w/ it's port exposed as 3306
Pretty sure you'll have to either:
- Connect HA to both host and bridge
- Expose ports to the host
You can expose a different port...
Tried option #1 last night and it's not possible. Confirmed w/ @raw hedge . Saw on StackOverflow that others were having issues with this.
Wish Docker team supported the ability to add a container to both host & bridge networking.
I think I'm going to have to go with option #2.
I'll give it a shot.
If I'm exposing / publishing ports from some of my containers to the host, then do I need to still set up service-level bridge networks between containers?
For example... I have a service-level bridge network created for my mosquitto MQTT container and zigbee2mqtt container.
Wait...figured out the answer. Because I'm using Docker Compose, a default bridge network is created between containers.
So I don't need the additional service-level networks, unless I want to isolate certain containers from others—but doesn't really matter since I'm exposing (publishing) ports on them.
@humble mirage So now that I'm going the HA host route, when I am trying to reference the other containers/services (at their exposed ports), rather than using the container name (aka. hostname, which seems limited to the bridge network), should I rather use localhost now?
Probably the IP of the host rather than localhost
So for example, rather than mysql://user:pass@mariadb... use mysql://user:pass@localhost...
Although, won't this require me to open that port in my firewall on the host?
I was trying to prevent opening ports like that.
Dang it! "sqlalchemy.exc.OperationalError: (MySQLdb._exceptions.OperationalError) (2002, "Can't connect to MySQL server on '192.168.10.125' (115)")"
I also tried localhost and get this message: "sqlalchemy.exc.OperationalError: (MySQLdb._exceptions.OperationalError) (2002, "Can't connect to local MySQL server through socket '/run/mysqld/mysqld.sock' (2)")"
"sqlalchemy.exc.OperationalError: (MySQLdb._exceptions.OperationalError) (2002, "Can't connect to MySQL server on '127.0.0.1' (115)")"
Sounds like you've got some networking issues to chase
Possibly caused by your firewall rules 
I disabled my ufw firewall completely.
I'll send my config...one sec.
Please use a code share site to share code or logs, for example:
- https://dpaste.org/ (select YAML for the language)
- https://www.codepile.net/ (select YAML as the language)
- https://paste.debian.net/ (you guessed it, select YAML as the language)
Please don't use Pastebin, since it can randomly add spaces to the main view. Please also don't share text as images since it makes it harder for people to help you. Remember that others may have colour blindness, impaired vision, etc.
Have you tried connecting to the MQTT broker with MQTT Explorer on Windows?
Oh firewall
No wonder
Docker should punch through but….
Man, I’m at the point where I think k8s is easer than docker
Disable your firewall till you have it workinf, then you can open it and open ports as needed
At least you will know at that point that it’s not docker / ha etc and only your firewall
This seems like so complex/complicated for a simple setup though...unless it's really not.
Haha well, I'm making progress. With firewall completely disabled and with a comment from @humble mirage re: changing references to use 127.0.0.1 instead of localhost we're getting places.
I at least have my HA container now talking to my db container.
Running a simple curl -v command to the db service while bashed into my homeassistant container (via docker exec -it ...) is at least returning some stuff.
So I think the issue is either w/ my firewall or maybe even my db instance. I may have to "blow away" my db instance and recreate.
I didn’t realize /etc/hosts is magic
🙄😂
Although when I tried using localhost, I get this message: "sqlalchemy.exc.OperationalError: (MySQLdb._exceptions.OperationalError) (2002, "Can't connect to local MySQL server through socket '/run/mysqld/mysqld.sock' (2)")"
So looks like not specifying an IP address cause MariaDB to try running through sockets instead.
Yah, that’s docker being a pita and yah that’s sockets all right
Ugh...Docker
Use 127.0.0.1
super extra doubtful there's anything I can do about this, but I can't get haos to start. at all. I've flashed the ISO for a generic x84-64 platform onto this drive with balena etcher, I set it as the boot drive, and it just... sits there. a blinking underscore and nothing else. no error, no response, I can't type anything.
@wise robin Is UEFI enabled in the BIOS of this machine?
So @humble mirage I'm not sure I completely understand what these components are or how I integrate them with each other. Did you simply freestyle it or did you follow a guide somewhere? I'm a little out of my comfort zone here 😅
Why does docker swarm suck?
I've been freestyling unix type operating systems for decades 
Key thing is to realise that remote access is a thing made up of separate steps
Start with DuckDNS - which gives you a hostname that resolves to your WAN IP. Once you've got that working, move on to SWAG.
well i have duckdns working
The people who created those Docker images have their own Discord (linked from https://www.linuxserver.io/support) where you can get help with those
that is I can connect to home assistant via my duckdns subdomain - but only using http
Right, so next step is SSL
You have various options, but SWAG is one of the better options
Did you read the intro for those images?
yes
It explains what they are, and what they give you
maybe i should read them again 😅
SWAG gives you:
- A (reverse) proxy server to handle incoming traffic
- SSL certificates
Issues, limitations, failed projects from them. They bought a company to offer persistent replicated storage for docker swarm then gave up and did nothing with it. You can’t connect devices to docker swarm containers without doing docker in docker…. Etc
Thx!
Only thing left from docker is hub, dev, and compose. They failed the cluster area but are fantastic for single server / self hosted
Makes sense 🙂
Everyone uses them to build their images for containers, so docker is not going away any time soon
i cant help it anymore , i need to ask 😛 Trying to get HassIO to run on a SSD M2 drive in a Argon Case. Bootloader is okay seeing as it does try booting from usb but keeps saying GPT : no bootable partition . Tried flashing it with PI imager and also Etcher, but no luck
Hey guys
Sooo...
here's where I'm stuck
installed HA UEFI to a "overkill" M.2 via balenaEtcher
then booted and went through its setup, system reset
now it sits at a screen that says autoboot
boot0 it will not continue on.. system freezes / kernel panic and then reboots
Quick question - how do I reset the passowrd for HA on the Raspberry Pi?
Found my answer in a forum post on zigbee:
To change a device path of an existing configuration, simply edit the file /config/.storage/core.config_entries
Save and reboot!
Does anyone know how to update a docker compose installation in portainer?
I want to update my home assistant version
nvm got it
I am installing HA on a RPI 3 and now after waiting around an hour i still am stuck on this screen
any idea's what this could be?
i flashed my RPI 3 16gb sd card with the rpi imager with the available HA image on there. i have a small router connected to my laptop and RPI without ethernet connection.
Click the blue dot
what i understand from that is that it needs connection to the internet?
Well, yes
Without that it can't download the required Docker images
(and then the required Python libraries)
Once the install is done you can cut it off from the Internet, though I believe the Supervisor may sulk a bit
alright thank you 🙂 The youtube guide i was following said i didn't need ethernet connection that is why i didn't get that sorted
the biggest issue is keeping up with time in HAOS as far as I've seen
Well, it doesn't need to be wired networking, but ... really you'd be daft to use WiFi
Also, YouTube guides are often crap
that is why i searched for this server to get some hands on experience advice 🙂
@midnight elk are you planning on using wifi? Or are you trying to go no internet with the pi
i will be using wifi when the setup is complete since i will need to use MQTT for sending data from energy sensors ( circutor cvm-1d and powerlogic pm5300)
Why do you need wifi for Mqtt specifically
i have now given it network access now it gives this error
@midnight elk Please use a code share site to share code or logs, for example:
- https://dpaste.org/ (select YAML for the language)
- https://www.codepile.net/ (select YAML as the language)
- https://paste.debian.net/ (you guessed it, select YAML as the language)
Please don't use Pastebin, since it can randomly add spaces to the main view. Please also don't share text as images since it makes it harder for people to help you. Remember that others may have colour blindness, impaired vision, etc.
(Applies to these errors if possible)
Dns error it looks like. See after a reboot
atm i am not able to get in the .local anymore after reboot gives me a refused to connect now
Use the IP
i don't know the ip of the PI itself i tried googling for a default ip and that one doesn't work either
sorry that i am quite slow on getting this work
Well, it's assigned by your DHCP server, most likely the router
You can check the router's DHCP client list to see if a Pi shows up, or use tools like nmap or fing to find it
There’s no default IP for it lol
after rebooting once again i can get into it with the .local again. but the nmap program does list 1 host but it says the host is down and it doesn't display a specific ip
hey, does anyone know how to try and recover a virtualbox installation gone wrong? I'm panicking as i think ive lost all my setup
Do you not have backups?
im not sure, i dont think so

You can or can’t use the .local?
also what nmap command did you run?
You might also be able to view it on your router
You only make that mistake once (hopefully)
i thought i had, but now cant find anything
What's the problem?
virtualbox went wrong and i wouldnt detect my bridged network adaptor. In trying to fix this I think i might have lost my HA
So you only lost the network adaptor? Or does HA not boot)
*?
Also are you running HAOS then
If it's just the network you can recover this
Ive reinstalled virtuabox and im not sure whats left of HA. Yes it was HAOS
ended up with it just booting to shell
As in the HA shell?
removed that, just made it worse and worse
the VM
Well, if you removed the VM, you're fucked
So you broke it yourself from a perfectly recoverable problem
Can I recommend that the next time you've got problems, you don't take the nuclear option?
Wait when you deleted the vm, did you delete the files (and if so check recycle bin asap) or just remove it from virtual box
no i didnt remove the files, I got a second VM booting up from the original files but I'm not able to connect to it still, so I'm not sure If it contains my configuration or just a virgin HA installation
If it boots does it go to the HA command line
Like the “available at homeassistant.local” one
You can connect using the console
Ignore the network problem for now - log in at the console
Also to confirm, you have a network adaptor added correct
at the moment the only way i could get it to boot was to chose NAT in virtualbox
Well what if you did network info on that command line? See what IP it gives. Otherwise follow tink’s idea
not sure how to do that
ok just trying that now
Did you do it
the command works but not having any luck trying to connect
What IP did it show?
^
One for your LAN, or something else?
Because, if you're using NAT, you can't reach it from any other computer
it got the docker ip and the othger ip which i assume is the ip for HA?
Re-enabling bridged mode networking is the way to go
but its not in the range of my network
Then that’s an issue
maybe because its on NAT
Follow what tink said
i'll try that again but there seems to be a problem with it
Make sure you pick the interface that's connected to your network
I've seen a fair few people screw that up
There's also often issues with WiFi, if your Windows PC is using WiFi
it is using wifi yeah, it been fine for months but i had some updates recently think its screwed it up
Can you use Ethernet in any other way?
i can run a cable
Server stuff is just a lot better being cabled tbh
still getting this - Nonexistent host networking interface, name 'ASUS PCE-N53 300Mbps 11n dual band Wireless PCI-E card' (VERR_INTERNAL_ERROR).
yeah i did that, reinstalled wifi card, ended up reinstalling virtualbox, finally removed the VM , which i really regret now!
ok just rebooted again, ran a cable selected the ethernet instead. just waiting for it to boot again
Thanks so much, its still there.. Thank god for that
now to run a backup immediately
And ensure that it's not left on that VM, so that it's actually a backup
yeah cheers, think thats where I went wrong actually
sorry for the late response, but it uhh.... might not have UEFI? I saw no option from it in the BIOS and this machine is an OEM desktop from 2008
I was able to get homeassistant running after installing proxmox a few days ago, but it was struggling to handle a VM (and also, despite Virtualization being enabled in bios, proxmox couldn't actually tell and fought me on the VM)
the bios itself is fully accessible, and boot diagnostics screen can identify all the hardware without issue
to specify, the machine was originally an HP KJ385AA-ABA a6433w, but some of the internal hardware has been changed
Installing proxmox on this same machine?
that's what I had initially tried. proxmox with Hassos as a VM, but it yelled at me for virtualization stuff. it worked, but very poorly because hardware virtualization had to be disabled
use one of the other instalaltion methods then? or different hardware
Hello,
I want to change from my ethernet connection to a wifi connection for HA, as I am having some problems with the powerline ethernet adapter. I enabled WiFi on the RPi3B+, but HomeAssistant has no connection to the internet. AS I have HA running in docker I assume it's a problem of the internet not being properly bridged to the docker containers. Anyone has an idea how to solve this?
The Docker container should be running in host mode
If it is, then it's going to use the working network connection
The other thing to do is to check https://my.home-assistant.io/redirect/network/ and see that HA is set to Auto configure
It says my HA instance is not supporting redirects.
Ah well, you can navigate to the Network view manually
https://my.home-assistant.io/faq/ is worth setting up
I tried restarting the homeassistant container with --net=host, but unfortunately that doesn't seem to work. I struggle to wrap my head around the docker install (I installed it a while ago and my knowledge is quite limited). So I don't understand if I have to start other containers as well with --net=hosted or how the other containers are started (by the homeassistant container?)
What does your docker command line look like for HA?
IMO you're a lot better off using Compose files, though if you're not planning on learning Docker you may want to switch to using HAOS
Would love to learn docker better but hard to understand it with just following the HA install, but yeah that's my bad.
What does your docker command line look like for HA?
What do you mean by that?
for homeassistant:
docker run ghcr.io/home-assistant/raspberrypi3-homeassistant:2021.11.3
That's it?
No wonder you've got problems
docker run -d \
--name homeassistant \
--privileged \
--restart=unless-stopped \
-e TZ=MY_TIME_ZONE \
-v /PATH_TO_YOUR_CONFIG:/config \
--network=host \
ghcr.io/home-assistant/home-assistant:stable
That's what the docs show
Well tbh I followed the install (not sure where exactly) and never bothered with anything else.
You're not even mapping /config so all your config is lost... like tears in the rain
You didn't follow anything official
Or you were drunk
I'd suggest you start fresh, with a compose file (see https://www.home-assistant.io/installation/linux#docker-compose), since your config is lost anyway
Then... sort out backups
And 2021.11... HA, like most software, does not age like fine wine
My config is not lost. It's inside my homeassistant docker container in a ´config´folder. But I don't remember the exact install instructions I followed.
So I agree, fresh install would probably be better.
Well, the wrong ones 🙂
Hopefully you have enough Linux-fu to be able to extract the config from inside the container
Thanks anyways, will put a fresh install on my to-do list.
should i install ha-os in a Proxmox VM (with my zigbee dongles passed through) or on a dedicated RPi-4?
VM > Pi
The Pi is a great educational tool
the thing is,
since zigbee has not so good range and my server cant be in my room i think RPi will be the only option
Buy a network connected Zigbee coordinator
Zigbee is also a mesh network, point to point range doesn't have to be "good"
I do not plan on running tensorflow on haos so i think performance will not be a problem
It's not just about performance
what else?
SD cards die
no sd card
Then it'll be ok
good
i will have a sufficient psu + UPS
Is it possible to overclock a RPi running HA-Os
?
No
why?
Because HAOS
Then ... don't use HAOS
would it be possible to implement this in haos?
i may look into this and maybe maybe submit a pr
debug ssh.... yes
for most users... no
ok
pretty sure HA OS is already pushing it with all the psu issues that come up
I'm always up for seeing yet another person with a broken install due to self inflicted power issues
||Why yes, I am a little evil||
||but I neeeed to run crysis||
I received a persistent notification notifying me that my storage was running low. I, ignorantly, assumed it was a bug and, without even thinking of checking the storage settings, restarted the Home Assistant instance. I am no longer able to access my instance (due to not having enough hard drive space). How would I go about accessing my Home Assistant Blue's (ODROID-N2+) internal storage so I can clear space even though I am unable to connect to it?
@unique crest please pick an avatar - it makes it much easier to notice when you post, rather than it being assumed your message is a continuation of the previous person posting
That said, you should be able to connect a keyboard and monitor
Thank you!
I have installed HA-OS as a Proxmox VM, and when it initially was created it allocated 70GB. I want to reduce the VM disk size. I have reduced the (data) partition to 30GB (using gparted) and that all works fine. I can boot back up. But when I reduce the ZFS disk from 70GB to 32GB (to avoid cutting into the last part of the data partition) it fails to boot. It shows me the options: autoboot, boot 0, boot 1 and shell. Only the Shell option works. If I increase the ZFS disk back to 70GB it works again.
Does anybody know how I can get the VM disksize reduced?
Could it be linked to how the overlay partition is mounted? Or maybe something with the EFI boot section?
@uneven panther honestly, just create another VM and restore a backup
VM's are like tissue paper
Hello. I have a Pi4 and I was trying to install home assistant supervised, but it complains about not having /etc/default/grub
It also says Device "eth0 <newline here> eth0" does not exist
Don’t use home assistant supervised, use the os or go the container / core route
Why not?
Core is missing too many features. And the OS means that my Pi4 is single-purpose.
So is supervised if you follow the requirements
I guess that I could install it as a VM. I got a case with a touchscreen for it, and it's a bummer not to really use it.
Urgent: I changed my carrier a couple days back and since then I can't connect to HA on 4G from android. When I am wifi works and when i give hotspot to other device works too. Other family members still using HA on 4G. Even from the browser of my mobile it's not working!
What method are you using for external access?
duckdns
Assuming you did not change your router, does your new carrier use CGNAT (a lot of cellular carriers do). That is incompatible with DuckDNS. To check have a look at the IP address reported on the DuckDNS website for your domain. Is it different from your actual public IP address reported by your router? Is your router's public IP address in a non-routable private IP range?
I am not a developer so not sure about the question but to be clear I didn't change my router or anything to do with my set up. I just switched cell carriers and even though I am still on the same phone I can't access my HA when I am outside of my home! My wife can still access HA from outside of home!
What about other Wi-Fi networks?
Your grocery store, Starbucks etc
It has nothing to be with being a "developer." You switched cell carriers. So you need to find out if the new carrier uses CGNAT because this is incompatible with DuckDNS.
Also this is an #add-ons-archived question.
is there a way to repair the supervisor if you cannot access add-on's anymore?
I know ha supervisor repair exists, but I cannot open the SSH web interface
already a long time broken since upgrading to Home Assistant 2022.2.9
and doing the Supervisor update at that time
Home Assistant Supervisor
error unknown
I powered it on and off 2 times, doesn't change it
it still works, but I'm stuck in 2022.2.9 now
so I hope there is a way to repair the supervisor, I use a RPi4 with Hass OS image
@royal sentinel maybe,
ha su repair
ha core rebuild
ha host reboot
I gave myself hotspot from a different mobile and it worked. And gave the other mobile hotspot from my phone it did work! But when I am on my freaking 4G it doesn't
You need a vpn or duckdns
And this is why Docker is superior.
They are using duckDNS. @whole coral have you done the tests I asked you to do yet?
Go to your router, find the public IP address. Go to your DuckDNS service web admin page (not the addon, the page you used to register the domain) and compare the address it reports. Is it the same?
@royal sentinel connected a keyboard and monitor to the pi?
I got supervised to work perfectly well on raspberry pi os in supervised mode.
Ducks a cnat issue
And I have a k8s cluster working perfectly on Raspberry pi os
Just because you can, does not mean you should
Then please help me understand why supervised is not a good way to run home assistant?
how do I uninstall the stable version to install the superior version?
Supervised requires you to be an expert in docker, networking, and Linux. OS does not
There’s functionally no difference or advantage between the two
Superior version?
That’s not the terminology for any of that
how can I add addons then?
If you’re running container, then you can start any container (which is what add ons are) otherwise you can install OS and dedicate the machine to that
The server is running a few more things so I cant install the OS
Then you can just install other containers and they’re add ons
All the add ons have container versions
It just requires some manual work to set it up
well then
I am having issues to connect HA with Amazon Alexa
(without subscription) following this video https://www.youtube.com/watch?v=Ww2LI59IQ0A&ab_channel=EverythingSmartHome
What part are you having issues with
6:23
If you run supervised, you may as well docker compose
I cant find the button to edit the configuration
im on the stable version
he has much more options than I do
Edit button on what page?
The dashboards?
It's an add-on most likely
but I dont have the add-on button in settings menu
No need to use addons etc
but I want the add-on button as well
is this already setup with DHCP? so it can get an IP address automatically?
Don’t set manual for ipv4 then
I want to have the add-ons button
do I need the supervised version for it?
I set it for auto. And it didn't do anything
Then I would wait for some one who know more about ha os
I need help installing Home Assistant over again but I would like to have the add-ons button in the settings
which edition do I need to install
previously I installed stable version for docker (and it didn't have the add-ons button)
@opal pebble why do you need the vscode add on. Just use VSCode on your computer and ssh in
Or run code-server
Then yes you’ll need to dedicate the machine to HA
MUST?
Supervised or OS both require only add ons and the core to run. Nothing else
So yes. Must
whats the dedicated virtual macine one
Especially since you can’t give a reason why you need it. Except for “I want it”
You run a virtual machine on the server and run HAOS in that
Its just easy to access the files in the directory and edit them
Aint it going to take MUCH more resource of the server?
Again, code server (for a web based interface) or just VSCode on your computer. They can do that. VSCode has a ssh extension that lets you work remotely
Yes. Because you want the extra stuff
Yes, it will use more resources. You are the one who wants to make your life more complicated
What you have will work fine right now
why it says its possible without OS than
Also sorry not going to open links from random (possibly scam) domains
Sorry it’s not one I know. Please can you use a more known image host
Because supervised is the same as OS in that you must dedicate your install to it if you want it to be supported
so without that I must use each addon as a container?
Yes. Using docker compose will make your life easy too
whats that
Google it
Way to do docker without needing command line arguments
Make your life easy
how can I get started with that one
Google is your friend
oof
Docker Compose version v2.5.0
Wait so now what do I need to do?
Docker compose: https://docs.docker.com/compose/gettingstarted/
Home assistant (scroll down to docker compose): https://www.home-assistant.io/installation/linux
Then ask more information and others may be able to assist (I’m about to sleep so can’t)
Though if you’re running a home server you should really get to know how to use docker & docker compose well
Yeah start learning it well. It will be useful
Hi anyone know how i can install full home assistant on Ubuntu 20.04
?
The raw.githubusercontent.com .... do not work
depends on what you mean by "full home assistant", but your best option is a VM with one of the HA OS images
or a container install
you're describing a Supervised install, and you can't do that on Ubuntu
and you really don't want to do it anyway
What will be the best way to install it on a NUC with ssd ?
one of the ways I suggested above, or you can install the generic x86 image directly
KVM on Ubuntu
Dose not eat resource?
🤷
It is installed on a pc or ?
how can I add a SSL certificate to HA website
openssl req -sha256 -addext "subjectAltName = IP:192.168.1.189" -newkey rsa:4096 -nodes -keyout privkey.pem -x509 -days 730 -out fullchain.pem
will this do the job?
yes, a Lenovo M20 tiny
Hi all! Spent some time last night trying to resurrect an old raspberry pi2b (I'd upgrade to a 4 if I could get my hands on one). I tried the 8.0 build for the 2 (at least I think it was for the 2), tried the 7.2 and the 4.x too but all of them fail to start up any docker containers. Should it work? Are those builds really for a 2? 🤔
You happy with an untrusted ssl cert?
as long as it says its secure
I have done this and restarted the container
but the certificate is still not secure
could you please have a look https://agamsol.xyz:8123/
Well you got a cert issued to a local ip and that’s not where you’re accessing it from, plus it’s untrusted too
If you want external access look at running HA through a reverse proxy. Like traefik, caddy, ngnix, etc
more information please
its also unsecure from the local addresshttps://imgur.com/AhT7HaS
Yeah untrusted cert. If you’re trying for remote access on your domain then look into a reverse proxy
Before you expose your HA to the internet I recommend you do read up on how a reverse proxy works and how to do it
Probably good for #general-archived now too
I'd like to add a SSL certificate to HA
using sudo ./letsencrypt-auto certonly --standalone -d agamsol.xyz -d agamsol.xyz
can I specify the port?
anyways I cant use ssl as ports 80 and 443 are taken by pterodactyl
I've been using linux since nearly the beginning (1997-1998), and I work with Docker every day at work when I'm not writing software. So I'll be quite all right. 😉
Afternoon folks, any insight into moving home assistant (with Conbee stick using zigbee2mqtt) from VM to a dedicated host? I assume i just need to install haOS, restore from snapshot, plugin the stick and make sure zigbee2mqtt knows it's dev address... anything else I need to know?
Use linuxserv.io swag container or traefik for a reverse proxy + TLS certs from letsencrypt
Or cert-manager + traefik on k8s
Example of k8s is https://github.com/miwagner1/k8s-at-home-with-k3s for swag, linuxserv has lots of docs on how to do it with docker compose
pretty much just that
and also if you have a static IP or a reserved IP that is port forwarded to update that IP into the new instnce of HA
or change the forwarding rule
Yeah, sits behind proxy. Should be good there. Much appreciated
also prefer to make a dummy account first and upload snapshot after rather than restoring during onboarding
nothing should be different but sometimes people exhibit some issues
Oh, that’s interesting. Good call
but you'll presumably have a keyboard and monitor plugged in so you can just monitor supervisor logs of what's going on if it comes to it
That’s the plan. Got a small dell 3020 and will just use the nuc image. It’s a new set up anyways. So I can afford down/tinker time.
Anyone had issues after brun the HSO on the NUC ssd to not boot up?
Hello guys! Is it me or the "homeassistant" pypi package isn't published since the date is used as version?
$ pip3 install homeassistant==
Collecting homeassistant==
Could not find a version that satisfies the requirement homeassistant== (from versions: 0.7.0rc1, 0.7.0, 0.7.0.1, 0.7.1, 0.7.2, 0.7.3, 0.7.4, 0.7.5, 0.7.6, 0.7.7, 0.8.0, 0.9.0, 0.9.1, 0.10.0, 0.10.1, 0.11.0, 0.11.1, 0.12.0, 0.13.0, 0.13.1, 0.14.0, 0.14.1, 0.14.2, 0.15.0, 0.16.0, 0.16.1, 0.17.0, 0.17.1, 0.17.2, 0.17.3, 0.18.2, 0.19, 0.19.1, 0.19.2, 0.19.3, 0.19.4, [...], , 0.96.3, 0.96.4, 0.96.5)
No matching distribution found for homeassistant==
Answering my own question, it seems to be here: https://pypi.org/simple/homeassistant/
My pip installation is probably broken
Alright, it's not shown in the list because pip seems to filter the output based on the Python version currently executed
Thanks again for your help! I just wanted to let you know that after troubleshooting, I found out that the reason I ran out of space was due to a file called "home-assistant.log.1". It was taking 110 gb of space! I had recently made some changes to the logger configuration so I suspect it was due to that. In one week my disk usage went from 2 gb to 110+ gb!
@unique crest that's the HA log from the previous session
You still have to dedicate the install to supervised too, so it’s up to you on whether you want that
why not just run HA in container (docker/podman/containerd) without supervised on host or whatever ?
That’s what I do
or simple python3 install behind nginx on baremetal
i wonder why it is considered legacy by many integrations/projects (or dropped altogether) to have yaml configuration
Ever since the new way of updating the update notification/entity doesn't show me an update button to click.. meaning I have to update by using the service call.
Does anyone have an idea what might be causing the button to not show up?
https://i.imgur.com/t8TkWNV.png
Yeah, i have seen that. It does not justify dropping yaml in most of the integrations
Ok
It is like ‘we will drop a way for you to automate so you can only go click in web gui’
And no keeping history for you
I just don’t get it
That page actually said that both will move forward but next thing happens - yml configuration gets dropped by lots of projects
Hi guys, would like to know the following. I'm planning to move my installation from a raspberry to an Intel NUC. Having some issues with my current running environment and it should be related to the SD Card but want to move to Intel NUC while I'm at it. Now one thing when restoring HA that I don't see mentioned anywhere: I expect that I need to restore to the exact same version of HA right? I didn't update for a while, running HAOS 7.6, there's a new Core version available. I didn't want to touch the Raspberry more than needed. Can I find the same older version so that I can install this on the NUC before restoring, or is the preferred way to first upgrade everything on the Raspberry to the latest version?
7.6 is the latest stable version of HA OS.
The how far behind the latest core version (2022.5.4) are you?
2022.3.3
But restoring a full backup will also restore the HA version it was made in
but you can also update to a specific version of Core via the command line
Ok well then it shouldn't be a problem if I install the latest version on the NUC, upload the full backup and just restore? That would be great
You don't have the choice really in the beginning anyway. The HAOS images don't have Core in them. They download the latest version upon first installation
but yes then when you try to restore a backup you should see under Core that it's the older version you'll be restoring
great thanks
Yes it's the same
Hi,
I'm trying to install HA, I'm now stuck at the Preparing Home Assistant screen for over 3 hours or so. And in the Network tab I see this error log constantly. '21-10-12 19:04:55 INFO (SyncWorker_2) [supervisor.docker.interface] Updating image ghcr.io/home-assistant/raspberrypi3-64-homeassistant:landingpage to ghcr.io/home-assistant/raspberrypi3-64-homeassistant:2022.5.4
21-10-12 19:04:55 INFO (SyncWorker_2) [supervisor.docker.interface] Downloading docker image ghcr.io/home-assistant/raspberrypi3-64-homeassistant with tag 2022.5.4.
21-10-12 19:04:56 ERROR (SyncWorker_2) [supervisor.docker.interface] Can't install ghcr.io/home-assistant/raspberrypi3-64-homeassistant:2022.5.4: 404 Client Error for http+docker://localhost/v1.41/images/ghcr.io/home-assistant/raspberrypi3-64-homeassistant:2022.5.4/json: Not Found ("no such image: ghcr.io/home-assistant/raspberrypi3-64-homeassistant:2022.5.4: No such image: ghcr.io/home-assistant/raspberrypi3-64-homeassistant:2022.5.4")
21-10-12 19:04:56 WARNING (MainThread) [supervisor.homeassistant.core] Error on Home Assistant installation. Retry in 30sec'
And I get a 404 error because it can't find the manifest.json file
Do I need to change that in my router settings or? Don't really know what you mean sorry
This did not fix it for me
Hello,
I am trying to install Home-Assistant with Supervisor in my Synology NAS using VM and I got into a couple of issues. (I am using the OVA image downloaded from https://www.home-assistant.io/installation/alternative )
- IPv4 is disabled and I only got an IPv6 (I solved it by configuring a manual IP and gateway in the CLI and disabling the IPv6 but could not set a nameserver)
- After finally starting Home-Assistant, I am not getting the supervisor option
Has anyone installed successfully using this method? Thank you...
Oh! thanks... I was not aware of that change... It´s been a long time since I used it supervised
It still cant find the manifest file.. Any idea else on how to fix this?
Change dns reboot router, once it’s up reboot ha os
The issue you are having is a network most likely dns issue
https://ibb.co/7t4ydck is this normal?
Is this the correct channel for SSL related questions?
Probably
So, I've been trying to setul SSL, and my https://x.duckdns.org:8123 url is working fine, but on my internal network:
- My homeassistant.local:8123
- I have to use https://ip:8123
- Mobile app doesn't work
In the Settings/General/Network i have a notice:
"You have configured an HTTPS certificate in Home Assistant. This means that your internal URL needs to be set to a domain covered by the certficate." So does this mean I have to generate a certificate also for the local url?
That's entirely normal and expected
If you want to use homeassistant.local or the LAN IP then you need to have a (reverse) proxy handle SSL
Thats a router option right?
(Or traefik)
Your external connection then uses that instead of going directly to HA
Ok so I don't need that, but when trying to setup the internal address via IP I get:
The certificate for this server is invalid. You might be connecting to a server pretending to be "X", which could put your confidential information at risk.
What can I do about this?
You ... use a reverse proxy server
Right now you've got the connection to HA protected by an SSL certificate for yourhost.example.com - if you use anything else to connect to it then the client quite reasonably refuses to connect
This is exactly what's supposed to happen
To avoid it, you stop having HA handle the SSL, and instead have all your external connections route through a reverse proxy that handles SSL
Alternatively can also just not use the IP and always use the domain instead
not that setting up a reverse proxy is complicated
Got it to work! I had to reinstall NGINX Home Assistant SSL proxy plugin
Ever since I changed from HAOS on Raspberry Pi to the VMWare ESXi install on Intel hardware, I'm having problems with my networking. The ESXi version just doesn't register the homeassistant name with my local DNS like the RPi install did. This is the first time I've used VMWare and I can't see what is wrong with the install or config. Is there someone who could help me please?
Thank you guys for the help! 🙂
Ah so you were using a reverse proxy all along 😛
This did not work, still getting this error. Even reflashed to older versions but no luck
Trying to upgrade to HA OS 8.0, running on VirtualBox, keeps getting stuck on "A start job is running for Docker Application Container Engine", doe snot progress during this for one of the components, runs over 15 min
Hi all, i'm new and trying to setup Home Assistant on a Raspi. At this moment i see the 'Preparing Home Assistant' for more then 20 miniutes. Anyone who can help me?
Click the blue circle
Thnx, I see a lot of data...
Please use a code share site to share code or logs, for example:
- https://dpaste.org/ (select YAML for the language)
- https://www.codepile.net/ (select YAML as the language)
- https://paste.debian.net/ (you guessed it, select YAML as the language)
Please don't use Pastebin, since it can randomly add spaces to the main view. Please also don't share text as images since it makes it harder for people to help you. Remember that others may have colour blindness, impaired vision, etc.
Only thing I can see is
22-05-16 14:17:26 WARNING (MainThread) [supervisor.host.network] Requested to update interface eth0 which does not exist or is disabled.
You did connect it to a network cable?
How can I disable windows updates on my ha machine. Driving me nuts
Here , try this:
http://letmegooglethat.com/?q=disable+windows+update
I just tried to update to 8.0, now it can't boot. Getting "incorrect compressed chunk" no matter if I select slot A or slot B, how do I get this fixed? I am running the Generic x86-64 image on a Lenovo Tiny
That only really shows me the windows sanctioned solution which isn't really a solution and windows registry hacks from random website.com, which I'd rather not follow without someone else having successfully done it before 😅
@knotty dagger I would open an issue on github on the OS repository
I can see Stefan has been active in the last 30mins already in a few issues for 8.0 so you'll probably have better chances there
Is proxmox basically an os that almost exclusively runs vms? That's what's looking like
Pretty much
Anyone running haos on a raspberry pi 2 or is that a no no? I tried and failed, but my docker knowledge on rpi2 is minimal so I'm not sure why it wouldn't work.
it's not recommended
Is there a trick to getting it working? I don't mind if it's just slow as it's just for tinkering (I can't buy a pi4 right now)
It should work, but the install process will take longer
Everything will take longer in general
Errors with not being able to find the cli. Docker ps shows nothing.
Is it connected via ethernet?
It is
hi guys, my HA was running fine for over a year, i updated recently and today i wanted to change something but i get
resolving "homeassistant.local:8123""/" ("") for "10.238.2.28" err: rpc error: code = Unknown desc = NXDOMAIN
homeassistant.local will only work if you're using Home Assistant OS, or you've named the host homeassistant
i have HA OS 7.6 like in the screenshot
server is booted
https://postimg.cc/dZ3Q9drm
this is what i get
https://postimg.cc/56HNYBxf
Then I take it that HAOS is running in a VM?
exactly
What if you connect to these:
If both of those fail I'd assume that an Ubuntu update has screwed you over
Try SSHing to the IP, the username will be root (from memory)
If that fails too then your problem relates to the Ubuntu host
ok i am in via ssh
How'd you do it? I updated to 8.0 and now it won't even pull an IP from my router.
They're not on 8.0
Ah.
Is anyone having issues after 8.0? I'm running it on a NUC
ha core logs will allow you to see what's going on with Core
its doing something, also my doorbell rang with an automation , but i cant access the GUI
Looks like HA is running but it's having networking issues
That again points to an Ubuntu update
ha supervisor logs
Just the core container that's having issues then ... weird
Check the Ubuntu firewall settings
i already restarted the VM = same
the ubuntu worked for over a year , i changed nothing there
i have auto. updates
Well, there you go
but i have a message, that some aren't done
Which says that some were
you are right, but yesterday HA worked 😦
So, since yesterday you had an Ubuntu update that's causing you issues
looks like
First thing to try is a whole host reboot
if you say so
Well, you lose nothing by doing it, and it may fix the problem
i'll give it a try, ty
-> Addons