2025.1: Backing Up into 2025!

[arapaimas]

Hi,
I saw nothing in the release saying about ways to modify a card by editing it, but today, after the update, when i want to edit something on my dashboard i can’t.
I had a Yamaha network player that disapered completely and states of ram/cpu of my unifi AP’s are unavailable.
You should stop “improving” on thigs that are working well more more than 1 year.
I understand that things are evolving but you can’t jusr remove things because whatever.
Many pages are not working, inlcuding system: Error while loading page system.
We live in these houses.

[Tryfos]

Please see my above post 227 in relation to a (very old) custom Samsung integration. Could they be connected? Unfortunately I can’t use any other integration, since my 2015 pre-Tizen HU8500 is only supported by this integration.

[RKor (Rani)]

Same here, broke my ecowitt integration “Config flow could not be loaded: {“message”:“Invalid handler specified”}”

Find a solution yet?

[Tekwyzard]

• how often do you open your backup archives?
  Quite often, especially since discovering some years ago that one could, and thus access the individual files contained therein

• What do you open it up for?
  Usually to see how the heck I did something however long ago, or to undo something I’ve just messed up, like tweaking an automation, config file or whatever, then something doesn’t work properly, and I was an idiot with a goldfish grade memory; so into the most recent backup file I go, to copy and paste what worked back into my running configuration. Much quicker, and less destructive to data collected since the last backup, than messing about doing a full restore.

• How often do you need to restore your backups?
  Very often, especially recently, when so many basic things that used to work so reliably are now often broken in early releases. I’ve given up installing x.x.0 releases for this very reason, as I was spending so much time restoring what previously worked.

[Ildar_Gabdullin]

@Recte

Card-Mod changes might be effected by this update

Correct, checking commits in frontend now & trying to find what could cause this.
This thread is occupied by a new shiny backup feature, so no room for discussing other things (much less critical for sure).

[CaptTom (Tom)]

@Edwin_D

I’m not 100% sure, but previously if you added a network storage folder in settings and pick that as default backup storage, I think it is created there right away. At least I saw a file being opened there as soon as I start a backup. And there also was no local file.

This would be a change. The way I understood it to work was like this:

Feature Request: Backup Directly to External Storage

You would be correct in your assumption. The backup process uses the Linux TAR (short for “Tape ARchive”) command to take the contents of your config folder and database and consolidate into a single archive file. The TAR process, by its nature builds the file locally in a temp directory, and then will copy the completed file to its destination. Likewise, if it was using GZip compression (which at least looking at my backups HA is not), then the TAR process would take a second pass at the file t…

This may be a moot point for many, but in some implementations the amount of writes, and available storage, can be issues. It’s always best to understand how things really work, beyond what the UI is showing you.

[KevinCathcart]

Yeah, I agree.

My concerns with this new system include:

• Lack of “automatic” partial backups right before updates. This has saved me before.

• Lack of per location backup retention settings. Right now the default behavior looks to be that all enabled locations will have the same set of backups, unless an agent includes special logic to retain fewer, or the user manually deletes backups from underlying storage. If that is true, this feels screwy, because there is a very good chance the user may want different retention locally vs some cloud based agent.

  I supose a user can configure to retain all backups, and manually create automations to delete from different backends, but the system provides no real help with this. There is not even services to call get a list of backups for a specific agent, or delete from a specific agent. Indeed, there isnt even a an api or webservice call for deleteing from just one agent.

• Lack of long term retention flexibility. I want to be able to have several days of daily backups, several weeks of weekly backups, and like a year of monthly backups. This is easy with the google drive addon, but would require complex manual retention automations.

• Insufficent documentation about manual decryption.

[Recte]

I don’t think it is up to you to determine what this thread is about, other than the fact that you are the one filling the thread with irrelevant posts. Mine is informative for those who read the thread before upgrading, yours is redundant at best.

[Piregwan]

Meross integration seems not working since this update (humidity, temperature, MS310, MS210…)

I just used backups and Meross integration works again. So Meross Cloud ioT broken

[CordoWEB]

If encrypted backup is a very good idea, force encryption without decryption tool is not usable !

• What’s in the backup : I do not know
• How extract just one file ? It is not possible
• How verity the content of a backup : It is not possible

Without decryption tool or the choice to not encrypt backups, it is not a good idea to upgrade to this version

[lobolobo]

It´s your integration the HA native one?

[gjgeary (GregGeary)]

I agree with per location backup retention settings.
Cloud is set at 1 - that’s OK
System set with a small number due to limited system storage
NAS set with a much higher number or longer period of time

Also the option for long term retention. These are standard features with many backup programs.

[dumbdevice]

Another complain about forced encrypted backups!

I hope to see in next release option to disable encryption for local backups. I don’t care about cloud backups.

[Bohemian]

I’m glad that you flagged this. I have several climate-related automations that would immediately break because of this change. This (and encrypted backups) gives me another reason to sit this upgrade out for now.

I understand some of the rationale for this type of thing, but ‘minor’ renaming / case changing of attributes / service calls is the number one reason I’m upgrade hesitant and the primary driver of reverting / restoring from backups.

I fully agree that this should have the proper public warning period first.

[Lateadapter]

Reading all that I’m very much worried regarding all the upgrades, which will come…
I’m using HomeAssistant two years now - it’s an awesome project and there are awesome people making all that happen. But some time I’m at least surprised …
I learnt a lot within recent two years - very fast I learnt not to hurry with updates, but starting very carefully reading all the messages. Today it did save me from surprises regarding the move to Zigbee2MQTT 2.0 but luckely the suggested solution in case of problems was already possible to find in the release notes.
However the upgrade to HA Core 2025.1 seems to become a real challenge. Not only the heavily discussed backup strategy (well intended but not so well introduced), but all the other surprises with custom components are scaring. I know, that there is a risk especially with HACS, but on other side many things till today are only possible using such solutions.
As a sample: I made a dedicated weather-dashboard for a wall mounted tablet, heavily using custom packages (custom:canvas-gauge-card; custom:config-template-card; custom:button-card; card_mod …)
There wouldn’t be any replacement in case of upgrade-caused problems. But how long such loved solution can stay at life? (I’m not a programmer!)

My lessons learned from today:

• I’ll skip the 2025.1 update for now
• beside the productive system I’ll setup a second system just for testing new updates (should be easy, because I’m on proxmox)
• I have to setup even a third HA instance, which could provide just the shown dashboard w/o updates “for ever”

[pove]

You just need to create a 3 lines script/automation, backup action is still working, they’ve only added more functionality to automate backups using UI and providing the basics for future improvements.

3…2…1… Backup

So I ran a couple of tests to see if there was a way: In one test, I used the “Backup Now”, which unfortunately still leaves the archive encrypted. The second test, I used the Service/Action “Home Assistant Supervisor: Create a full backup” action: hassio.backup_full data: name: Test Unencrypted2 This archive is unencrypted and the individual files can be browsed/opened.

On the other hand, and as a Nabu Casa subscriptor, I have to thank about the free space for cloud backup. They’re are adding benefits for the same price, you have to like them…

Maybe what they have to do is to silently announce new versions, until it’s stable for a full month or something like that, many people complain about things they don’t even have (because I assume they didn’t install the latest version).

Or just leave it as is, forum gets animated…

[akuehner (Andy)]

MY HACS integration broke from this update. Not just integrations from HACS. After reloading the HACS integration that started working again (earlier reboots didn’t bring it back), but now HACS shows zero downloaded/installed integrations. I have a lot, and some of them still seem to exist within HA itself, but HACS doesn’t see them. One that was completely broken was mushroom cards, which I now reinstalled.

If I go to HACS it thinks HACS and Mushroom cards are the only integrations I’ve installed through there. That’s way off.

[dumbdevice]

@pove

You just need to create a 3 lines script/automation, backup action is still working, they’ve only added more functionality to automate backups using UI and providing the basics for future improvements.

Well, I have to use a workaround now instead of perfectly working good old manual backups from the UI.
I just gonna wait for the next release, I still have my old backups from the last year around
I think it was a big mistake to force encrypted backups.

[lmagyar1973]

@tom_l

3. Add-on and core updates no longer have the option to backup before updating which could lead to up to 24 hours data loss depending on the time of update vs the time the last automatic daily backup was made.

All issues raised during beta testing. All ignored / dismissed.

OMG, I’ve planned to open an issue, but this is intentional???

Can’t beleive anybody on Earth thought this is a good idea.

[septillion (Timo)]

I’m glad to see I’m not the only one who thinks it is ridiculous to have encrypted backups without decryption tool. It really clashes with the open mindset of HA.

Don’t get me wrong, I’m all for encrypted backups. Even if it was for local. But a big (the biggest?) protection of the backup is from myself! It is when I make an edit to a dashboard or automation that I should not have done. And for that I want to restore that (or retrieve a part) and not a complete HA install including database en LTS.

So yeayyy encryption. But booooooooooo locking the files away…