Right, but what I’m saying is, the 2 of you are following a path that doesn’t make sense because the 2025.x.x version of HA has no impact on running an update. What matters is the supervisor and the OS version you’re running as both of those technically have a hand in the update process.
Meaning, updating to 2025.4.3 will not solve your problem. Fixing supervisor will likely fix your issue. Updating HAOS may also fix your issue.
My issue was not a supervisor issue that worked fine. My issue was a completely unresponsive UI and not automations working. I guess it could have been supervisor related, but SSH digging (the little I had time for) did not seem to indicate that.
Here is my path → Upgrade to 2025.4.1 (or 2) from the completely functional UI. Update applied, PI4 reboots. I see in the web UI everything restarting and the final “Started” message. Try to use anything in the web UI - NOPE. Hope automations still trigger - NOPE. Access via SSH things seem to work.
I will hopefully have more time next week to dig further into my issue, it appears the other is resolved so they might have only appear to be related.
May give HAOS a try, but there seem to be a fair number of issues being reported there as well so a bit spooked.
Have you tried updating to the latest RC version of Bermuda (if you don’t have it already)? Version v0.8.0-rc1 is available from HACS, and works (for me) with HA 2024.4.2.
Thanks @karwosts for the new feature, its definitely very useful even as it is.
Just being able to see total untracked consumption in the daily total is personally all I’d like to see. The breakdown of the parent into child and untracked is nice, but if (like me) you have a parent that covers a significant portion of the house energy and then a lot of individual devices below that, it could be cluttered quite easily.
I have both WiFi and POE BT trackers that stop reporting tracked device distance when Core is updated. I’m on latest HA, Bermuda and ESPhome (including new BLE updates in 2035.4).
Not sure where to report this, but since a couple of versions my energy cost calculations go haywire whenever there is a core update.
I have anomalies on April 7th, 13th, 15th, 18th and 20th. Note that I might have installed them a day after release, so these dates don’t necessarily match the release date.
The costs of external energy usage as well as the compensation for adding power to the net looks like this:
Any idea what’s going on here? I’m pretty sure it is linked to the core updates alone.
I know that I can correct these more or less, but it’d be nice for it not to go haywire on each update.
Thank you and best!
Edit:
It’s probably noteworthy that I’m using integral sensor helpers to accumulate the power usage across my Shellys, as I don’t have a main sensor for my flat yet.
Then another two template sensors to get the current power draw from the net as well as returning power to the net. Something like this:
You should add an availibility on such templates. Otherwise you will temporarily have wrong numbers if one of them is not available and because of that 0 in your case.
My Aqara FP1 sensors seem to have become unreliable since late last week, not updating their presence and/or missing leave events. At first I thought it was just one sensor that might have been picking up some new vibration/movement and tried repositioning it several times, but the situation did not improve. Today I noticed another sensor doing the same thing prompting me to examine the history graphs in HA and realised that ALL the FP1 sensors I own have had their occupancy states stuck on since the 25th, which I think may correspond to when I updated to 2025.4.4 (from 2025.4.0).
ZHA update in 2025.4.3 seems suspect. Restoring a backup from a week ago and will monitor to see if the issue persists… Edit: Initial results seem promising, sensors are no longer stuck on since restoring the backup, but I need to keep monitoring. Edit: One of the sensors is still not firing the leave event and is stuck on, so it might be a different cause. I’ll keep monitoring / digging.