Needs help to understand Home Assistant

Hello everyone.

I’m new user with HA.
Today, I have already an application to manage my home. it’s Jeedom a french application.

I’m working to migrate on HA but I need to help to understand the nuance.

I have a virtual machine with Zigbee2Mqtt.
My HA is a different virtual machine (provide by OVA Image)

In HA, I added MQTT plugin and I configured my broker.
In Zigbee2Mqtt, I activated the home assistant option.

After this all my device are added in HA, it’s ok so far.

often, I noted the device are unavailable.
example with door sensor. It is the same with a plug or swith

I need to understand why I have this behavior.

Thank you for your help

The best way to deploy for your particular circumstances is the spin-up a KVM or better an LXC with Mosquitto (i.e. have the MQTT broker on its own VM instance).

Then your Z2M talk to Mosquitto as well as HA. It could be the HA behaviour might have to do with the VM resources in your HA.

My vm HA has 2GB of memory.
I will test with more

my mosquitto is already in external vm because it’s used by other system

Are the devices which become unavailable all battery-powered? If so, they may be sleeping. In your example the contact sensor seems to be waking again when the door is closed, so it may be nothing to worry about.

it’s the with wire device, it’s strange.
i’m going to test with available option on zigbee2mqtt

I’m seeing this the same way with my shelly device

image

That means that the device stops talking to HA (or in this case Zigbee2MQTT).

What make and model of sensor? What are your availability settings in Zigbee2MQTT?

I speak this option.

but as above, shelly are impacted also so it’s not zigbee2mqtt but it’s seems that it’s my HA VM.

I added 1Go memory but already the same.

My HA has a bug with refresh of state.

Then it’s possibly network related - check the log files (Core, and Supervisor, and Host).

(unless the Shelly also uses MQTT, when it may be a broker problem)

I checked log and I have a lot of log like:

Logger: pychromecast.socket_client
Source: /usr/local/lib/python3.11/site-packages/pychromecast/socket_client.py:678
First occurred: 13:49:21 (567 occurrences)
Last logged: 17:03:20

[Cinéma(192.168.1.185):8009] Heartbeat timeout, resetting connection
[Cuisine(192.168.1.186):8009] Heartbeat timeout, resetting connection
[Maison(192.168.1.186):32001] Heartbeat timeout, resetting connection
[SHIELD(192.168.1.183):8009] Heartbeat timeout, resetting connection
[Chambre(192.168.1.47):8009] Heartbeat timeout, resetting connection

Some of those is normal, I get a few hundred of those a day for each entity.

Ok so I will test HA during few days.
Thanks You

Hello,

I rebuild my VM from OVA HA official Image.
Before to import other devices, I’m going to test with just Yeelight.

Normally, everything must be ok without Delay

A little post to indicate that I reinstall HA on Docker Platform and it’s seem more stable.
So the issue is OVA template.