I recently migrated from Odroid N2+ (HA Blue) to an Odroid M1 due to the availability of M.2 data storage in the M1, in addition to the eMMC. My original HA instance had HACS and 3 apps, as well as voice assistance, and several add-ons. On the M1 I downloaded and installed the latest image and proceeded to install HACS, and the rest of the add-ons and integrations I had in the N2+. Then I saved a complete backup from the N2+ and imported/restored it into the M1.
All seems to be working fine, but I now have two “devices” for each of Whisper, Piper, and openwakeword under the Wyoming Protocol Integration, whereas the old N2+ system had only one each, with entity names such as “tt.piper” and “tt.piper_2” for the second instance. If I delete the second entity, it soon shows up again, asking me to configure it, and accept it. I have not tried to delete the non “_2” entries, as I am afraid that woudl totally mess things up.
I am wondering if this was caused because I went ahead and installed Piper, Whisper, etc, on the new system BEFORE I did the restore from backup ? which apparently installs inegrations/apps it sees are needed ?
What woudl have happened if I had not installed HACS, and restored the backup ? woudl the restore process have installed HACS and the apps I was using (like Alarmo?)
What woudl be the best parh forrward to fix the voice app duplication ? Turn off voice support, delete all bits, and start over ?
It is likely all my fault for loading add-ons and HACS prior to restoring my backup. However this is not clearly stated in the official “hardware migration” instructions. Further, HACS being not being an official part of Home Assistant, I figured that those backed up settings for my HACS apps would need HACS and the apps to be installed prior to restoring the system using the back up.
However, the issues I have are only with official “Add Ons”, nothing wrong with any of the HACS apps, or MQTT, or ESPHome – only voice apps, and HomeKit Bridge which are all duplicated.
Any help for a noob woudl be greatly appreciated, as I do nto want to break things further by just poking it without knowing.
You clearly state that you installed all your add-ons and integrations, you had on N2, so why don’t you move on with your thoughts, and install latest image on M1( No add-on, no integrations), and then restore with your Full backup from N2.
What could go wrong ?, you still have your full back-up of your N2, And i assume a working N2
So, just to make sure I understand, you are saying I should start all over again, with a fresh install of just the OS image onto the M1, and then use the N2 backup on the M1 to restore my system ? What happens to HACS ? will it get installed ? …though maybe since I have no issues with any of the HACS bits, perhaps I do the same and isntall before backup restore…
Unfortunately, the N2 system got wiped out, as I am reusing the N2 eMMC memory card for the M1.
I suppose my other option is to fix the double instances in voice assistance, by deleting all voice assist add-ons and settings, and then re-install the voice-assist, and then do same for the HomeKit bits…
I will post again with results, currently away from town until Friday.
If you look at Developer Tools → States, the original entity will be “unavailable”, and the _2 entity will show the current state of the entity.
The only way I am aware of to remove the double-naming is to delete and reinstall the integration, then restart Home Assistant to discover all of the existing devices.
Learned a few things. I reinstalled the image for the M1, and did NOT load up any add-ons, or HACS. I restored the system using the full backup I had from before, and all came back ! including HACS (which is not officially part of HA…?)
The double instances of the voice-assist add-ons did not show, that is until ESPHome had an update. Then a RaspiAudio Muse Luxe speaker that I had been using fine earlier seemed to require the additional “_2” instances of Whisper, etc… so I deleted the install of the RaspiAudio and the prompts to install additional instances of Whisper, etc. are gone. I have not re-installed the RaspiAudio into ESPHome yet, but will.
Anyways, it is good to know, that as stated in the Documentation, a system migration can happen by just using a full backup after installing the base system image – which is exactly what the instructions state. However I got a bit confused about whether HACS and its apps woudl get restored, since HACS is not “officially” part of HA. Perhaps the migration blurb can state something to the effect that even HACS will be restored from a full backup on migration.
It’s not officially part of HA in the sense that HA is not responsible for anything HACS related.
However, that doesn’t mean that any integrations you install via HACS, including HACS itself, aren’t included in a backup. Once it’s installed, a custom integration is still an integration in HA’s eyes, therefore it’s included in your backups.
HACS is Home Assistant Community Store. It is a collection of community developed and supported stuff and not supported by the core developers. The files are copied and restored with backup and restore. Just like any other config file.
With an upgrade, there is NO assurance that any custom thing has been tested with the upgraded core or HAOS. Testing and any fixes needed by the custom thing is entirely the responsibility of the custom developer. Most developers do a good job staying on top of pending releases.
I say thing because you can have custom integrations, components,. automations or cards, so I call them collective custom things. And, not all from HACS. Before HACS we had to download things from each custom developer’s Github. I still have one or two things that never made it into HACS. What a pain.
If you are using any HACS things, you should periodically open the HACS control panel to check for updates. Often if a core or HAOS upgrade “breaks” a custom thing, the custom developer has already addressed it with an upgrade.
Thanks again for the tips. I guess I took this message below from the logs too seriously, and made me think HACS was not an official part of HA – am good now !.
WARNING (SyncWorker_0) [homeassistant.loader] We found a custom integration hacs which has not been tested by Home Assistant. This component might cause stability problems, be sure to disable it if you experience issues with Home Assistant
Though struglling now trying to figure out why I get the error below in the logs, and what it actually is… it’s happenieng over and over, I must have about a 100 of them, always the same –
2024-04-13 12:38:47.301 ERROR (MainThread) [aiohttp.server] Error handling request
Traceback (most recent call last):
File “/usr/local/lib/python3.12/site-packages/aiohttp/web_protocol.py”, line 350, in data_received
messages, upgraded, tail = self._request_parser.feed_data(data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File “aiohttp/_http_parser.pyx”, line 557, in aiohttp._http_parser.HttpParser.feed_data
aiohttp.http_exceptions.BadStatusLine: 400, message:
Invalid method encountered:
b'\x16\x03\x01\x02'
^
2024-04-13 12:38:47.329 ERROR (MainThread) [aiohttp.server] Error handling request
Traceback (most recent call last):
File “/usr/local/lib/python3.12/site-packages/aiohttp/web_protocol.py”, line 350, in data_received
messages, upgraded, tail = self._request_parser.feed_data(data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File “aiohttp/_http_parser.pyx”, line 557, in aiohttp._http_parser.HttpParser.feed_data
aiohttp.http_exceptions.BadStatusLine: 400, message:
Invalid method encountered:
b'\x16\x03\x01'
^
I’ll do some digging to see if I can figure it out, but I may have to come back and ask for help from the experts