Good news and Bad!!

[AndrewTapp]

Good news and Bad!!

First for the good news

I'm going to release v2.3.4 later today.

This release fixes the issues around entity ids and should now work properly. The original issue was around entity ids being incorrectly removed when the integration was removed. Please note that the history will remain.

The integration can be reconfigured without deleting and re-adding.

Re-authorisation if you should change your userid or password.

Recommended flow: remove, then re-add (with cleanup)

When you want a clean registry and preserved history (e.g. after an update that adds or improves async_remove_entry):

1. Update the integration via HACS so the new code (including cleanup) is installed.
2. Restart Home Assistant (optional but recommended).
3. Remove the integration from Home Assistant: Settings → Devices & services → Integrations → SolarEdge Optimizers → Delete. This runs async_remove_entry and cleans entity and device registries for that entry.
4. Restart Home Assistant (optional but recommended).
5. Re-add the integration with the same Site ID, username, password, and options. You get fresh entities and a clean registry; history reconnects because unique_ids are the same.

and now the Bad news

I've had some comments from people that the integration is not updating when there appear to be changes in the SolarEdge portal.

I've spent quite a bit of time investigating and can report the following.

You all know that the data within the SolarEdge portal is not real-time and can be delayed by up to 15 minutes, give or take.

If my integration finds stale data (over an hour old) then it sets the values to 0, however in my investigations I've found the following.

Bare in mind that for me sunset was around 17:27 GMT and therefore I shouldn't be producing shortly before this time and certianly not after.

At 17:41 GMT I took these screenshots, one from the SolarEdge app and the second from my HA optimizers. As you can see panel 1.1.1 all values are the same (please note that I have adjusted the time in my integration by +1:00 so it should read 16:30).

and this is the HA log entry

2026-02-12 17:41:12.361 DEBUG (ThreadPoolExecutor-9_0) [custom_components.solaredgeoptimizers.solaredgeoptimizers] Decoded JSON object for optimizer 154302877: {'countrySettings': {}, 'description': 'Panel 1.1.1', 'lastMeasurementDate': 'Thu Feb 12 16:30:42 GMT 2026', 'manufacturer': 'SunPower', 'measurements': {'Current [A]': '0.31', 'Optimizer Voltage [V]': '32.5', 'Power [W]': '18.60', 'Voltage [V]': '60.75'}, 'model': 'SPR-MAX3-400', 'phase1Measurements': {}, 'phase2Measurements': {}, 'phase3Measurements': {}, 'secondarySerialNumbers': {'serialNumber': '130FCCF8-E6'}, 'serialNumber': '130FCCF8-E6', 'softwareInformation': {}, 'status': 'Active', 'type': 'POWER_BOX'}

At 17:52 GMT I took these screenshots, one from the SolarEdge app and the second from my HA optimizers. As you can see panel 1.1.1 all values are the same (please note that I have adjusted the time in my integration by +1:00 so it should read 16:42).

and this is the HA log entry

2026-02-12 17:51:25.481 DEBUG (ThreadPoolExecutor-11_0) [custom_components.solaredgeoptimizers.solaredgeoptimizers] Decoded JSON object for optimizer 154302877: {'countrySettings': {}, 'description': 'Panel 1.1.1', 'lastMeasurementDate': 'Thu Feb 12 16:42:51 GMT 2026', 'manufacturer': 'SunPower', 'measurements': {'Current [A]': '0.5', 'Optimizer Voltage [V]': '45.75', 'Power [W]': '31.31', 'Voltage [V]': '62.62'}, 'model': 'SPR-MAX3-400', 'phase1Measurements': {}, 'phase2Measurements': {}, 'phase3Measurements': {}, 'secondarySerialNumbers': {'serialNumber': '130FCCF8-E6'}, 'serialNumber': '130FCCF8-E6', 'softwareInformation': {}, 'status': 'Active', 'type': 'POWER_BOX'}

At 17:55 GMT I took these screenshots, one from the SolarEdge app and the second from my HA optimizers. As you can see panel 1.1.1 all values are the same (please note that I have adjusted the time in my integration by +1:00 so it should read 16:47).

and this is the HA log entry

2026-02-12 17:55:30.121 DEBUG (ThreadPoolExecutor-12_0) [custom_components.solaredgeoptimizers.solaredgeoptimizers] Decoded JSON object for optimizer 154302877: {'countrySettings': {}, 'description': 'Panel 1.1.1', 'lastMeasurementDate': 'Thu Feb 12 16:47:15 GMT 2026', 'manufacturer': 'SunPower', 'measurements': {'Current [A]': '0.15', 'Optimizer Voltage [V]': '29.5', 'Power [W]': '9.23', 'Voltage [V]': '61.5'}, 'model': 'SPR-MAX3-400', 'phase1Measurements': {}, 'phase2Measurements': {}, 'phase3Measurements': {}, 'secondarySerialNumbers': {'serialNumber': '130FCCF8-E6'}, 'serialNumber': '130FCCF8-E6', 'softwareInformation': {}, 'status': 'Active', 'type': 'POWER_BOX'}

The worrying thing from this (from the log entries) is that the SolarEdge portal is running around 60-70ish minutes behind and not just the 15 or so previously experienced.

In v2.3.4 I have amended the stale data check from 1 hour to 2 hours, so at least you should get data. Hopefully this is temporary on SolarEdge's side. Time will tell.

On a second issue I've had some comments that data such as lifetime energy isn't flowing into the integration.

It appears that SolarEdge are in the process of migrating to a new portal setup called "digital-twin". As with the original the optimizer data is undocumented at least outside of SolarEdge and is NOT intended for public consumption. Hopefully the old portal endpoints will remain and we'll still get data through this integration.

I intend to do some work over the coming weeks to try and get to the new endpoints for "digital-twin" and update this integration accordingly, no promises though!!

In the meantime if any of you find out any information on this please let me know.

[AndrewTapp]

Just for information I may have a way forward to use the new SolarEdge One portal, early days but there is light at the end of the tunnel!!

[kaijk]

Thank you again for all of your efforts on this.

[AndrewTapp]

Just an update for your information.

For me as the current SolarEdge portal gets progressively delayed data as the day goes on, at the start of the production day it's around the 15ish minutes we all know, but towards the end of the day for me it can be over an hour.

I've managed to get the SolarEdge One endpoints working on my integration and have been running for just under a day now and is all looking good. The frequency of updates is improved compared to the old portal, although it still follows the 15ish minute rule, sometimes more sometimes less. Currently for me it's around 5-10 minute updates.

I'm going to continue to test with these new endpoints and if all goes well will release a new version (v2.4.0) that lets you choose between old and new towards the end of next week, maybe Thursday or Friday.

As with the original SolarEdge portal these endpoints are not publicly available and are intended for internal use and as such could change at any time.

[AndrewTapp]

I've just released v2.4.0. You' need to update through HACS, restart HA, then remove the integration, restart HA, clear browser cache, and then re-add the integration.

[kaijk]

I'm getting data with 2.4.0.

However, with both 2.3;4 from 2.2.3 and now 2.4.0, I continue to get
entities like sensor.optimizer_1_2_1_solar_current_1_2_1 with a solar
prefix. (I also tried reusing and no prefixes.).

When adding the integration it continues to find the Area that I had added
way back. I think my Device and Entity registries are chock full of
iterations of this and maybe confusing HA's cleanup process.

Here are debug log data from adding and removing V2.3.4 (with a s_edge
prefix)
Adding:

2026-02-14 08:55:28.830 INFO (MainThread)
[homeassistant.helpers.entity_registry] Registered new
sensor.solaredgeoptimizers entity:
sensor.optimizer_1_1_1_s_edge_current_1_1_1
2026-02-14 08:55:28.831 DEBUG (MainThread)
[custom_components.solaredgeoptimizers] SolarEdge Optimizers: Platform
setup complete for entry 01KHEH821TX9XYBYKE4X8NBG62

Removing:

 2026-02-14 09:03:00.687 DEBUG (MainThread)
[custom_components.solaredgeoptimizers] SolarEdge Optimizers: Unloading
config entry 01KHEH821TX9XYBYKE4X8NBG62
2026-02-14 09:03:00.695 INFO (MainThread)
[custom_components.solaredgeoptimizers] SolarEdge Optimizers: Removed 205
entities and 34 devices for entry 01KHEH821TX9XYBYKE4X8NBG62
2026-02-14 09:03:00.696 DEBUG (MainThread)
[custom_components.solaredgeoptimizers] SolarEdge Optimizers: Closed API
sessions during unload
2026-02-14 09:03:00.696 DEBUG (MainThread)
[custom_components.solaredgeoptimizers] SolarEdge Optimizers: Unload
complete for entry 01KHEH821TX9XYBYKE4X8NBG62 (unload_ok=True)
``
and some of what's left in `core.entity_registry:

{"aliases":[],"area_id":null,"categories":{},"config_entry_id":null,"config_subentry_id":null,"created_at":"2026-02-14T16:55:28.016758+00:00","device_class":null,"disabled_by":null,"disabled_by_undefined":false,"entity_id":"sensor.inverter_1_s_edge_power_1","hidden_by":null,"hidden_by_undefined":false,"icon":null,"id":"52616a41c2de92778227eed5e81a382a","labels":[],"modified_at":"2026-02-14T17:03:00.694579+00:00","name":null,"options":{"sensor":{"suggested_display_precision":0},"conversation":{"should_expose":false}},"options_undefined":false,"orphaned_timestamp":1771088580.6970067,"platform":"solaredgeoptimizers","unique_id":"01KHEH821TX9XYBYKE4X8NBG62_s_edge_power_1"},

(I had saved these a few days ago and had left the integration deleted
until 2.4.0 today)

[AndrewTapp]

Hiya, I will take a bit more of a look, but to be honest it's got me stumped at the moment. It's due to entities not being deleted properly, likely from a previous setup. All the testing I've done on my own setup works with no issues. It's early days but I’ve not had any similar comments.

As I said I will take another look but it's not looking good at the moment.

[kaijk]

Do you know enough about how the Device and Entity registries are built to say that I could likely delete all references to solaredgeoptimizers devices and entities with grep -v solaredgeoptimizers against the device and entity registry files? Then, start over clean. I'd make copies first, of course.

[AndrewTapp]

I've had some dealings with them and can do a bit of research tomorrow and give you my thoughts if that's good. Definitely take copies just in case, and HA backups.

[AndrewTapp]

Apparently the files to look at are as follows.

/config/.storage/core.entity_registry
/config/.storage/core.device_registry
/config/.storage/core.config_entries

But be very careful!!

I would suggest deleting the integration first and restarting HA before doing anything with these files.

The historic data remains and if an integration is setup with the same entities HA will automatically relink to this data.

[kaijk]

I see an entry in core.restore_state as well. Here goes...

[kaijk]

I backed up HA, stopped the Docker Container, and made copies of the registry files. Then I ran grep -v solaredgeoptimizers on them, overwriting the originals. This resulted in HA labeling the edited device and entity registry files "corrupted" when I started the HA container. After copying back the saved files HA started up OK.

[AndrewTapp]

Closing this, referencing #14