<ctrl>F5
to clear cache and refresh
That did it. Thanks
I agree, god this thread is soo negative… (some have relevant comments though)
Good that they have started implementing this, works for me as expected and advertised. I will still use Google Drive Backup in parallel for my files and also copy them to a local server until they get all the features in place.
if it’s mandated to be encrypted, I’m out….
Skipping this one…
I noticed this evening that all of my sensors are missing data from ~4:45AM to about 11AM for today only. At around 11AM, all of my energy consumption sensors reported crazy negative numbers, which completely messed up my energy dashboard since there was no data for hours and then sensors that should never go negative were now reporting negative values.
The only changes I’ve made to environment this week was the upgrade to 2025.1 and enabled this new backup system (that fails to upload to cloud due to my backup size).
I recognize this is not a support thread, but I honestly do not even know where to begin in preparing any sort of bug filing on this. It just seems super coincidental that at the time this new backup feature starts, the recorder stopped recording sensor data…and then added significant negative numbers when recording resumed.
Since discovering this, I’ve disabled the new backup feature (I was previously using OneDrive), manually changed all of the negative values to zero at 11AM today, I’ve added several exclusions to my recorder configuration (in attempt to lower db size), and purged/repacked db. But please let me know if anyone has any pointers on what information or logs to grab to file a bug to the right place. This was a painful issue with no clear reason why or how.
Does anyone else have problems with automatic backups? A manual backup is successful.
The automatic backup could not be uploaded to the configured locations cloud.cloud.
Yeah same here!
Same here and for the second time.
Same here. I’ve seen this behavior from day one of the 2025.1 release. I suppose this is caused because everyone has the automatic backup set at a fixed time: 4:45.
full on disaster recovery is not the only function of backups. Partial restores to the file or file content level are part of backup function.
Version control is about assigning meaning and function to each version, it’s about allowing data to diverge and get merged back into meaningful version.
Can’t speak about others but in my case I need backup function not version control. A better term for what some of us are looking for is probably history.
now I will grant that there is a lot of gray between backup, version control and history which is made even worse my some modern implementations in popular programs (office 365).
I haven’t upgraded yet, but I have been following the discussion around backups.
One thing I’m still confused about, is if the local archives produced by the new automatic backup feature are indeed encrypted?
Enforcing encryption for backups that are uploaded to third-party cloud storage is reasonable in my opinion. I don’t think though, that the local archives should be encrypted as well.
My suggestion would be to make a two-step process.
- Create unencrypted archive on a schedule and manage those using retention policies.
- Only encrypt the archive before it’s uploaded to the third-party cloud storage.
This way advanced users would still be free to handle their data as they see fit. Users who don’t want to worry about this can enable Nabu Casa backups and enjoy a convenient and secure backup solution.
You can still do partial restores of the backup, nothing changed here. If you’re digging into the big backup zip file, looking for an old version of a yaml config and checking if it’s the one you need, you’ll enjoy version control.
Anyway, this conversation will end once they add a simple button to download an unencrypted version of the backup, or something like that. And they will.
But if on the route, they can give a thought (or add to roadmap) to a way to give users this kind of go back to a previous working version of an automation/script/esphome/config/…, these days feedback will have been worth it.
I have been using the daily automatic backup integration for several years.
I have currently rolled back from 2024.12.5 to an older core version (here: 2024.10.4), so far, so good.
Of course, I have to update the individual integrations and HACS step by step - but some require higher core versions, e.g. 2024.11.
How can I select a previous core version for installation? Via the GUI I only ever get 2025.1, which I explicitly want to omit.
Thanks for any advice or links.
are you using HAOS, if so you can update to any version you want via the Command Line
(if you search the forum for something like “ha core update --version” you’ll get loads of examples and explanation on how to do this)
Good point. I’ve always used the term “archival” backups to distinguish them from DR backups. The schedule and retention period are different for each, and it’s typical to save DR backups off-site and archivals locally. Typically archivals are incremental backups, while DRs are full. Some operating systems incorporate version control into their file systems (e.g.; “previous versions”) to give an automatic archive capability.
Just to reiterate, if you are using any custom solution for backups (blueprints, automations, or add-ons that do not use encryption), they will continue to work today. Even with everything new, we’ve made sure to keep everything backward compatible.
As always, we’ll keep improving the functionality of automatic backups over the coming releases. As mentioned in our recent livestream, we’re looking at ways to decrypt your local automatic backups from the UI. This would be helpful if you needed to access the files in them. Hopefully, we’ll have more on that soon.
Thank you for the clarification and for the wonderful hard work you all do for the system!
I plan to have a second system to restore a backup so I can have a way to decript the backup.
So my current automation at 3am will still run and be an unencrypted backup?
You are right on your more nuanced definitions.
History is the most accurate description which is part of what VC provide but the latter includes the whole distributed concurrent edits thing we don’t necessarily need here… Unless you are in a multi-tinkerer household