Same symptom on an Apple-based setup — with some Thread-level diagnostics
I’m seeing the same thing with a TIMMERFLOTTE, but on a different controller stack, so it may be useful to compare.
Setup: Apple Home is the primary Thread fabric (Apple TV as border router). Home Assistant runs separately on k3s with the matter.js server. In addition, I have an SLZB-MR3 running OTBR that I joined to the same Apple Thread network using the operational dataset, but configured as routereligible disable — so it sits in the mesh as a plain child and does nothing except let me run ot-ctl diagnostics from inside the network.
What I can measure while the device shows “No Response” in Apple Home:
meshdiag childtable: link quality 3, frame error ~2.3%, RSS -61 dBm,conn-timeseveral hours with no reattachdns browse _matter._tcpshows the device’s SRP registration is alivedns resolve <MAC>.default.service.arpareturns a valid OMR address with TTL 300ot-ctl pingto that address: 10/10 packets returned, RTT 575 ms to 14.5 s, all delivered in a single burst (the parent buffers until the sleepy child polls)
So the device is joined, registered, resolvable and reachable at IP level — while the controller reports it as unavailable. That puts the failure above the Thread and SRP layers.
I’ve also tried a battery pull on the TIMMERFLOTTE. It rejoins the mesh immediately and shows a healthy link in the child table, but Apple Home still reports “No Response”, so power cycling doesn’t help.
One data point that may matter: I had exactly the same behaviour with a MYGGBETT contact sensor and same pattern shows up on a KAJPLATS bulb. The last one mains powered and acts as a Thread router — it’s the parent for all my sleepy end devices and routes fine — yet it intermittently goes unavailable in Home as well, while ot-ctl ping to its OMR address keeps answering with 0% packet loss. I’ve also shared the KAJPLATS with Home Assistant (matter.js server). It’s on the same Thread network and the same radio link, but from Home Assistant I can turn the bulb on and off reliably while Apple Home still shows “No Response” for it. Since the device is controllable from one Matter controller while another reports it unavailable, this looks like a controller-side issue rather than a device or network problem.