Fritzbox DECT thermostats go offline at regular intervals

Third party integrations
1

Hi all,

any idea why all my thermostats go offline (“not available”) for 30 seconds regularly?

Bildschirmfoto 2025-10-14 um 11.31.59
Bildschirmfoto 2025-10-14 um 11.31.59565×706 30.4 KB

Bildschirmfoto 2025-10-14 um 11.32.09
Bildschirmfoto 2025-10-14 um 11.32.09565×706 32.8 KB

HA 2025.10.2 running in Docker (TrueNAS) - if that might be the cause of the problem I can switch to a HAOS VM.

2

Same thing with HAOS.

Bildschirmfoto 2025-10-16 um 08.55.51
Bildschirmfoto 2025-10-16 um 08.55.51565×319 16.2 KB

So it seems it is not the Home Assistant side that causes these outages. I’ll check with the AVM/Fritz forum.

Nonetheless - any insight or ideas would be greatly appreciated.

3

I had the same issue. It seems there were an (by me unnoticed) update of my FRITZ!Smart Thermo 302 to 5.31. After a reboot of my Fritzbox and re-intialization of the Fritz integrations everything works fine again.

4

@jhol-byte thanks.

But it’s all thermostats at the same second that are “lost”.

In the the meantime I enabled the debug log. This is what I get every time the issue occurs:

2025-10-18 15:56:42.827 INFO (SyncWorker_0) [pyfritzhome.fritzhome] Updating Devices ...
2025-10-18 15:56:42.828 DEBUG (SyncWorker_0) [pyfritzhome.fritzhome] self._sid:9182f48209c7a3f8
2025-10-18 15:56:42.836 ERROR (MainThread) [homeassistant.components.fritzbox] Error fetching 01J3V037TRJC17N1C0FW2738EX data: 
2025-10-18 15:56:42.836 DEBUG (MainThread) [homeassistant.components.fritzbox] Finished fetching 01J3V037TRJC17N1C0FW2738EX data in 0.009 seconds (success: False)

Any way to increase the debug verbosity and have it tell me why there is an error fetching the data?

My network is pretty professional and stable for a home lab. I am a network engineer by profession. Although I will definitely set up a continous ping and try to find the Fritzbox dropping its connection every once in a while next.

Anyway a reboot of the Fritzbox and reinit of the integration won’t hurt. And let’s see what this Mikrotik switch can log or SNMP trap about flapping ports or some such.

Kind regards,
Patrick

5

So 24 hours of ping every second. It’s definitely not the network. It must be the Fritzbox not answering the API calls every once in a while, because why would all DECT devices fail and recover at the exactly same time.

I will now let a tcpdump run over night and examine what I get tomorrow.