Hello, SmartLife Beta integration no longer works properly since version 2024.5.x
When Home Assistant restarts, we have to redo the configuration. Do you have a tip to avoid this inconvenience? Thank
The 2024.7 update has broken a bunch of my automations. What sucks is that when you now go to the automation you see nothing inside that would point you in the right direction to fix it.
It appears that many of the device ID’s have changed to something different. Can this not be changed to just alert you that there is an issue and not disable or wipe out what is inside the automation ? By stopping other components from running the automation it could cause potential major problems. For example if I have a water leak now my water leak automation would not run. Not good…
I don’t believe all the problems are caused by custom integrations. It appears there is an index missing in the DB that causes a full table scan during the purge, and that in turn can cause a deadlock.
For me, the purge just runs forever. I do not get an exception as some do with the problematic custom integrations.
3 Likes
Use entities instead of devices and it will save you work in the long term even though the developers have gone out of their way to make using entities in automations more longwinded, entities are easier to fix
1 Like
I’m also experiencing the same, lots of history on all sensors. I do not have any of the three custom integrations mentioned in another thread (one of them being iCloud), and so far haven’t found a log entry that mentions anything about the recorder, so I have no idea what’s causing the problem on mine so far.
History comes back after restarting, but stops at 4am (happened at 4am twice now since upgrading on Wednesday)
A little search goes a long way…it’s in this thread and the link below…if it’s not one of the custom integrations then disable the purge of your database…
2 Likes
If you know how to open your automations.yaml file outside of the HA interface you should be able to see the device references that don’t display in the editor
I always try to use entities when I can, but unfortunately some don’t work. For example in the pic below, the device has multiple triggers available
In the screen print below the entity only has in “From or To” the following:
Any state (ignoring attribute changes)
Unavailable
Unknown
Any suggestions ? Thanks in advance…
Use an event trigger instead of a state trigger.
The problem with that is, it has the old device ID’s in the automations. There is no way to know what device this actually is. Especially if you have hundreds of devices like I do…
In the UI?
No idea sorry, I don’t use it. Though that looks like it could be it.
The event trigger docs are here: Automation Trigger - Home Assistant
Great, thanks. Seems to be working so far, no loss of history at 4am this morning
Confirmed on my side, not related to custom integrations and recorder stop at 4am. If you remove the auto purge, don’t do a service call to purge, it will fail also 
I tried to recreate my beloved lawnmover dashboard from grid view on the sections view.
In my grid view setup I got 4 colums to display everything on my tablet (screenshot 1).
In the section-view I only get 3 (way too large) colums in one row (screenshot 2).
Of course I set 4 columns for this view.
It would be nice if you get the same amount of columns displayed as in the grid view when using the same space.
Seems the space between the columns is causing this problem.
I’m in the exact same boat. I currently have switches attached to picture-glance cards in multiple rooms, so this select thing would not only require me to redesign several dashboards, I’d have create a bunch of yaml template switches to get it working with google assistant as well as it does now. This change seems to be obtuse and purposefully hostile to the end user - whoever did this seems to be motivated by the old “well I don’t use this so no one should have it” approach to design.
The only saving grace was that I noticed this in the change log before I upgraded so I didn’t break everything.
4 Likes
Same here.
Are there any indicators how many users/installations are Victims and lost their Historical Data in return for high CPU Load?
While it is not worst Case, i think this still is a really serious Fault.
I am wondering if this Update shouldnt get pulled and/or reverted by a Hotfix.
4 Likes
My long standing zwave js addon has been broken by the release-20247 update. The notes say
- If you use the Z-Wave JS add-on, you need at least version 0.6.0.
My version is Z-Wave JS Current version: 0.5.0 and no update seems to be available even if I try to force it with the update entity.
The last lines of the log show:
2024-07-08T01:27:22.454Z CNTRLR [Node 014] The node is alive.
2024-07-08T01:27:22.454Z CNTRLR [Node 014] The node is ready to be used
2024-07-08T01:27:22.455Z CNTRLR « [Node 014] ping successful
2024-07-08T01:27:22.478Z CNTRLR [Node 011] The node is alive.
2024-07-08T01:27:22.478Z CNTRLR [Node 011] The node is ready to be used
2024-07-08T01:27:22.479Z CNTRLR « [Node 011] ping successful
New client
Client disconnected
Code 1000:
All of my entities have gone.
I have tried addon re-starts, HA re-starts, power cycled the USB zwave controller.
Anyone know what I need to do?
The latest on my system is 0.6.2