Problems with latest HA version

I updated to the latest HA version (2021.1.5) a few days ago. Ever since then I have to reboot my Pi at least 2 times a day because HA is not accessible via web browser. I also seem to have lost all of the previous snapshots I had saved. It seems as though every time I try and log into HA I have to reboot to get in. When it does finally boot up, response times from HA to a device is very slow. Is there any known issues or bugs in this latest release that could be causing this? Is it possible to downgrade to a previous version? Thanks

1 Like

It is possible to downgrade (not sure it is possible on all of the environment) … Did you check your power supply ? This is one of the main root cause for instability

I am using the power supply that came with the Rpi kit. It appears to be working fine. Every time I need to reboot, the Pi is powered up and network traffic lights show network activity.

Are you trying to log in with http://homeassistant.local:8123? If this is problematic you could try using the actual IP address eg http://192.168.1.99:8123

I get the same results if I use local IP and the Nubu Casa URL, on and off local network.

That’s far from the “latest” version, try 2021.2.3 instead :wink:

1 Like

That’s the version I’m on…I looked at the top of the HA page and saw 2021.1.5 listed and mistakenly put that as the version I updated to.

I lot of people have the same problem. I had to downgrade to OS 5.5 and delete HA database to fix my problem that’s similar to your. However it’s most probably database problem. It’s probably too big (around 900MB)

This is exactly what I had in mind as read through all this ha-audio desater. The ha gets more and more unstable, if you dont take the time every release and refactor everything. Or worse, doing a complete new setup, because the configuration found in the documentation some years ago, is no longer supported… wtf
Im really into HA and spend hours per day to optimize and improve. But even for me hardcore user, its getting more and more frustrating and not fun at all anymore.

When updating to 2021.2.X I found my database got corrupted, causing slow response. I stopped HA, deleted the database, restarted and all has been fine since.

It’s not ideal that the update has caused a corrupted database for us few users but it’s hardly a widespread issue with cause to call the project a joke.

2 Likes

what error messages are in your log file?

Excuse my harshness, but cool me once fair enough. Fool me three times, well… :sweat_smile:

Well… you did post two threads basically about the same issue…

That is unfair! There are a few issues but, unless you’ve contributed significantly to the project, you (like most users) really can’t complain.

This is an amazing platform and works incredibly well given the right hardware and support - and user care. Simple things like reading the ‘breaking changes’ section before upgrading make a big difference.

Rather than call it crap, maybe offer your help to fix the problems?

1 Like

I can’t remember the last time I had to change anything because of a release.
I think the issue is when people skip some releases then upgrade.

Agreed. My process is to read the update notice, run the release through the ‘Check Home Assistant configuration’ tool, Understand what is likely to need changing then do the upgrade. I’ve been running it for probably about 2 years now and can honestly say that any failures have been self-inflicted!

I have not even updated. It just crashed out of nowhere this time…

If it crashed what do the logs say? Is the hardware OK? SD card failure, perhaps? I suggest running it on an Intel NUC with an SSD - you won’t look back.

This appears to be the issue I am having now. I guess the db is malformed. So will need to figure out how to delete it via cli

@beauwilliams
I went searching the forums for this issue because since the upgrade my HA instance has become unstable and crashes at times during the night and evening. Logs shows an sql error i have not seen before, looks alot like posted above. My next move is to erase the database as well, just need to figure out how to do that the right way using the mariaDB addon.

My first thought was that my mSata disk on the proxmox host was going bad, but reading this thread i realise it may be some kind of bug corrupting the database that is the root cause.