Classic network entities and devices become randomly unavailable

Hi there, for about 1 month all my entities and devices that rely on “classic” network have been getting randomly unavailable together at the same time.
It happens at a random time, and z-wave or zigbee device are not affected, only the ones using MQTT or Tasmota. Neither MQTT or Tasmota services restart nor HA reboot solve the problem, only host reboot or manual turn off/on will do. At first I suspected an issue with the IP reservation process on my TP-Link Archer C7 router, but all the affected devices have static IP addresses, and they stay reachable via a navigator, only are they unavailable from inside HA.
I had a look at the HA core logs but found nothing related to this at the last issue timestamp. Is there a way I can access more detailed network logs ? Does anybody experienced the same issue ?
Thanks for your help :grinning_face_with_smiling_eyes:

Update : I found out that this happens each time there’s a new update available in settings > update
It’s like http transport towards network entities is broken after the updater successfully fetched a new update.