Upgrading from before 0.104 to 0.104.2 has fixed the groups so you don’t need to do anything
Yeap… can confirm… I finally restored configuration from snapshot back to 0.103.6 and upgraded to 0.104.2 without any modifications in registry and everything seems to be fine!
1 Like
I switched over from zigbee2mqtt as I wanted to move away from MQTT and I prefer the integration on the deCONZ add-in. I’m using a Conbee II and I notice that the range is significantly worse than before, to the extent that several of my sensors will not connect and others are unreliable.
I’ve tried switching channels, and using a router but this not improved it.
To be fair, my previous stick had an external antenna, which may account for the difference. Also, my house is quite large, with concrete walls and floors.
Has anyone any thoughts on improving the range or has anyone tried an external antenna on the Conbee II?
Thanks!
How many mains powered devices do you have and what brand/model? What are the end devices that cause problems? Could be that they don’t connect to nearby routers, Xiaomi ones are known for that.
Did you try putting the conbee on an usb 2.0 extension cable?
I had the duplicated groups problem. Then suddenly after another reboot it duplicated all the lights and devices. I just removed the Integration, rebooted and re-added it and it was all fixed.
Upgrade from 0.103.6 to 0.104.2 and everything seemed well.
Then I rebooted my Hassos VM because of unrelated NUC/Proxmox memory upgrade, and now HA cannot connect to deconz anymore:
Error connecting to deCONZ gateway at 172.30.33.3
Config entry for deconz not ready yet. Retrying in 5 seconds.
while in the phoscon UI the gateway and all sensors and lights are available.
Restore to 0.103.6 fixed it again.
Can anyone confirm that the 0.104.2 upgrade and deconz survives a hassos reboot?
Yes, it’s on an extension lead which did help a little. All my mains powered devices are Z-Wave, and I don’t want to change them as they are working fine. All my deCONZ devices are sensors (door, motion & water leak) as I like their price and response times - noticeably faster than Z-Wave.
I tried using a CC2531 with an external antenna as a router, but devices don’t seem to connect through it.
hey all,
im running a conbee2 with the Hassio Addon.
i can access it via the UI and my lights work
however the motion sensor doesnt (Philips Hue Presence sensor - state doesnt change from "detected" · Issue #2322 · dresden-elektronik/deconz-rest-plugin · GitHub)
can someone help me with how i can check what these are via the rest api please?
The sensor should send a message when it no longer detects motion, which the REST API plugin picks up to reset
state.presence. What is the sensor’sconfig.delayvalue?
Hi all!
My Aqara switches are driving me CRAZY! Please, I need some expert help:
System:
- HASSIO Latest Version on Rasberry Pi 4
- deCONZ Latest Version + Conbee II Latest Version Firmware
- Devices connected via deConz: several lamps (various models), 1 Osram Motion sensor per room, and 1 Aqara WXKG03LM (single rocker) switch per room, plus an additional Aqara WXKG02LM (double rocker) in the kitchen (controls 2 devices in the back yard).
The Aqara switches are connected to their corresponding light automations through the deconz_event trigger, which then run some scripts.
Simple:
- alias: Living Room Switch ON-OFF
description: ''
trigger:
- platform: event
event_type: deconz_event
event_data:
event: 1002
entity_id: sw_living_room
condition: []
action:
- service_template: >
{% if is_state('light.living_bulb_light_1', 'on') %}
script.livingroom_lights_off
{% elif is_state('input_boolean.livingroom_scenes_auto', 'on') %}
script.livingroom_auto_scene
{% else %}
script.livingroom_lights_on
{% endif %}
Now everything was running smoothly until I added some additional Aqara switches. Now seemingly random switches fire random or even multiple automations to which they are not connected in anyway. For example, my backyard switch turns on the living room light, etc.
There is not even a similarity between two entity_id names, so, I don’t understand how that switch triggers that automation. It’s driving me crazy!
It seems as if the deconz_event trigger ignores the entity_id I put there and triggers no matter what.
Anyone had any similar experience? I need to understand if it’s deconz the problem, or my switches, or something else. Any ideas?
Thanks a lot!
This is my conf:
trigger:
- platform: event
event_type: deconz_event
event_data:
id: smart_switch_blabla
event: 1002
I’m using ‘id’ instead of ‘entity_id’
1 Like
I’ve had some issues in the past with duplicate groups, I was able to query the Rest Api with this info: https://dresden-elektronik.github.io/deconz-rest-doc/configuration/#aquireapikey
So basically first ssh into hassio, then from there I used curl, you can find my queries in this thread
I have upgraded to 104.2. Now All my deconz devices get unavailable every hour for 1 or 2 minutes. Has anyone had this behavior?
1 Like
There is no changes that should cause this. This is something with your environment or other integrations
I have had no other changes in my system. Never the less I will try to find the origin.
Yes, i’m seeing the same thing too. I’m running both home assistant & deconz in docker containers and have updated both , so I’m not sure whether it’s HA or deconz that’s causing it.
In the deCONZ logs i keep seeing this around the time of the drop out
20:03:22:865 Remove websocket 192.168.1.20:37826 after error Unknown error
I need to look through the HA logs too, but they’re pretty hard to follow using less or tail when debugging is enabled. The formatting is a nightmare as the log entries don’t seem to start on newlines (maybe just my setup?).
EDIT:
I’ve been following the logs and looking for some clues. The first occurrences of the devices going offline coincide with the time i updated to 104.2, so i’m going to rollback to 103.2 to see if it mitigates the issue.
I use IKEA repeaters and they work well!! (3 floors house). Suggestion with Xiaomi devices: make the association directly in the spot where the device will be used; it will connect to the local router during the association.
1 Like
I will also try the same. Please update the results and I will do the same.
One more thing: did you get the error in the log by using debug mode or just a standard error? I don’t have any errors other than errors in other component that use sensors from deconz.
The log message I’d pasted into my comment was from the deconz logs.
Jus checked deconz logs and have the same error
19:20:42:923 Remove websocket 192.168.1.77:40340 after error Unknown error
19:58:07:128 Remove websocket 192.168.1.77:45302 after error Unknown error
20:58:16:152 Remove websocket 192.168.1.77:46276 after error Unknown error
Don’t have a clue on what might be causing.
I can now confirm that errors on deconz started when I upgraded to 104.2 (I have been running deconz container for some weeks now without any restart, so I can only assume that the error has it’s origin on home assistant 104.2)
I’ve got deCONZ running in a docker, using ports 8080 (instead of 80) and 8443 (instead of 443). Home assistant configurator isn’t seeing it - is that because of the port change? Is there a way to direct home assistant to look for deCONZ on those custom ports?