Hello,
I’ve not a big knowledge with home assistant, I’m using it since approx 1 year on a rasp 4 with mainly zwave_js integration and some appdaemon scripts.
Since approx 2 months, I observe that zwave entities (zwave_js) are unavailable for approx 6 times a day, and then come back again with good behaviour.
The periodicity of these unavailable states is approx every 4h (but not always, it varies from 3h30 to 5h) and lasts just 40s.
This is annoying for me since sensors and scripts can have wrong behaviours during theses periods.
I’ve tried to investigate deeper, and I observed that cpu usage has peaks approx 95% when the zwave unavailability occurs.
Moreover, I observed the memory usage is growing from 50% (just after an unavailability period) to 90-95% when the zwave unavailability occurs.
The boot time is not impacted by these unavailability period (for me, system is not restarting).
The complementary modules installed do not show a big memory usage.
Could you give me tips to investigate deeper ?
I would be thinking about a memory leakage somewhere, but really not sure…
Below a zwave entity (switchCumulus1), boot time, and memory usage.
Home Assistant 2022.9.5
Supervisor 2022.09.1
Operating System 9.0
Interface utilisateur : 20220907.2 - latest
Many thanks for the help you could provide 