I only have one possible (PITA - pain in the ass) solution to try for this and not sure if it will work, so here I am.
No longer a noob, running Hopme Assistant Supervised on a RPI 4 w/8Gig or Ram on a 1TB SSD (not even using any microsd). Tweaked to high heaven and working perfectly, running very cool with low CPU etc. All automations working no issue. I run headless and only use telnet to access the underlying OS. Installation is both supported and healthy, OS and all HA integrations as well as HACS addons and integrations all up to date. No network changes.
Today I am unable to access the HA UI - from a web browser or the Android HA App. I also cast the UI to a Google Home Nest Hub Max (or whatever that hardware is called, got it free from my ISP for switching to them). That display gives me an error message about the HA websocket API or something or other (as it has before whever there is an issue with HA which up until now was rare). Up until now has been working pretty much flawlessly for some months. No errors in the syslog, all automations still working fine - I just cannnot get to the HA front end! I can telnet to the machine.
I have tried a complete reboot and shutdown - power off, restart - also I have gone through these normal steps for the HA Supervised installation from here:
(but not replacing trhe OS, just the end stepos which has fixed my issues before):
re-download and reinstall all the required libraries
re-download and reinstall docker
re-download and reinstall the OS-agent
re-download and reinstall homeassistant-supervised.deb
and the same thing - it comes up and automations all work fine, i just still l cannot access the HA UI by any means (but can still telnet to the HA Debian 11 OS command prompt).
I do have another SSD which is an exact clone of the SSD I have been using. It was created by simply booting up the RPI on a Raspbian microsd card, then plugging both SSD’s into it and using the SD Card Copier app. The backup SSD has always worked perfectly when I shut down and restart the RPI with that one plugged in instead.
However, today, HA has the same UI access issue when running on the backup SSD instead!
I would think Ok I now have a hardware issue. But, the display is headless and access is through the network, and everything else on the network is working perfectly as well (automations,. etc.). Lastly, I have weewx running on the host outside of HA, reporting my weather station data to many websites all over the world (and I am using the ambientweather integration as wel, which is working perfectly still too). And those websites are all getting timely updates from my weather station via weewx runing on the RPI as well. So it seems not to be a hardware issue if it is related to networking.
I am thinking I can do my next attempt (the PITA attempt mentioned in the beginning of this message) reinstall everything from scratch (the OS, etc.) on one of the SSD’s and then if I can get into HA I can easily restore the HA config from a backup - but will that work - if the backup SSD which was working fine does not work either?
I use duckdns for access and went to their site, that seems to be up…
HALP! (Pulling out hair) Thoughts before I try the drastic starting from scratch (which might not work as the backup ssd doesn’t work)?
Thoughts?