Hi,
I’m also thinking about moving deconz to “bare metal”.
Since you’ve already done this: is it possible to set up deconz on a new pi and then use a backup done within the HA add-on to have all the devices available again on the new pi? Or do I have to connect and set them up all again?
Yes that should be possible
Hi. Yes. You create a backup from the Phoscon App.
And on the new Deconz you upload the saved backup (same screen in Phoscon)
Thanks for the quick answer, will give it a shot the next days.
Hi, I have deleted integration, restarted HASSIO, now when I try to add deconz in integration it asks for HOST IP address, ,in the past it gave it automatically (173 … something) , now nothing. How to know which address to put?
@KennethLavrsen how do you run your Deconz in Ubuntu? Is it bare metal (i think so) or some VM? Frankly speaking I’m getting tired with Deconz addon to HA not being too stable and not reporting properly states to HA, so I’m considering swithing to solution that would not require restart of Zigbee network every time I restart HA. However I’m limited in possibilities, so I can install Ubuntu in ESXi, install Deconz in, passthrough the Conbee 2 stick to VM and give it a try, but once again it is virtualized env, with Zigbee gateway not connected directly to Deconz…
Currently I’m running HA as ESXi VM, which includes also all addons… I’m wondering if moving to separate VM for Zigbee will improve stability. It will remove Docker from the picture and will make Zigbee independent of HA restarts, but will it really help? Also what is your experience with Deconz releases? I think Dresden Electronics is releasing new versions more frequently that these arrive to HA. How it works in practice?
Just to make it clear, when you restart ha, deconz won‘t restart. All docker container should run, unless you restart your hassOs! This shouldn‘t be this often.
My setting here is working like a charm.
I have more then 80 zigbee devices (osram, Ikea, Xiaomi and hue), no problems with the smart plugs and all are reporting fine!
To find a solution for your problem, I have several questions:
- what is your Setting?
- how many devices do you have?
- how many routing devices (wired devices) do you have?
- on what machine is hassOs installed?
For example, my initial setting was a rasbpi 2 with some devices. As my setting was growing, my rasbpi needed over 40 minutes for an update. Also my new ikea blinds didn‘t work properly.
That was why I replaced my rasbpi to a nuc. No Linux, just hassOs directly on it!
Well updates just takes about 5 minutes, restarts are done in a couple of seconds and my zigbee setup, with deconz as an addon, runs like a charm!
Hi jan, i am currently running hass.io on vm on a nuc and facing many stability problems with the conbee stick. What are the disadvantages from your point of view when someone runs hassos directly on nuc? Is it difficult to maintain the containers?
Personally I’m running HA VMware image on ESXi server. For ZigBee I’m using Conbee II stick passthrough to HA VM. I’m running Deconz as addon within the same VM. Majority of my devices are permanently powered (38 Ikea bulbs, 2 Ikea LED drivers, 4 power outlets), only remotes (4 Ikea on/off switches) and motion sensors (4 Ikea and 8 Aqara) are battery powered. As powered devices are spread across entire house in quite uniform manner, I do not see need for repeaters.
I made some WiFi reconfiguration recently, to avoid signal interference and this improved stability of Zigbee network substantially. Issues that I’m facing are rather on HA-Deconz connection… not all devices are always visible to HA (even if these are visible in Deconz) or their state is properly reported (e.g. HA thinks it switched off all of bulbs, but some stay switched on). I think it might be partially also related to Deconz itself - for example when I switch off deconz group of light and one buld stays on it has nothing to do with HA (it switches off single entity and get state update from Deconz, that all devices are switched off). This should be Zigbee network issue, since devices in such case are located 0.5~1m away from each other.
After restart usually I quite fast see all devices being available to HA, but evry few restarts it happens that one or few devices are unavailable in HA, while Deconz see these. In such scenario running deconz.device_refresh sometimes help… but not always.
Also what I do not like in this setup is that from time to time Conbee stick stops to be visible to HA (i do not see it HW config), so also Deconz could not find it. In such case restart of entire ESXi host is required to make it available again. Those are not frequent cases, but make life less comfortable, especially that it happens usually after HA restart, that I’m doing quite regularly for updates of software or configuration changes requireing restart.
1 Like
I run deconz on bare metal on Ubuntu Mate LTS. Installed with apt after adding the Phoscon repo. And I run the beta branch as I am always after latest device support
In /etc/apt/sources.list.d/deconz.list
deb [arch=amd64] http://phoscon.de/apt/deconz focal-beta main
Then apt update and apt install deconz
See the official docu. It is so easy
1 Like
Do you have the stick in the nuc USB port? Because conbee sticks don’t like usb3 ports. Extend with a usb2 cable and the problem is solved.
Well, I run hassOS directly on a nuc. No container, no vm.
There is a whole thread about this
you will find some pro and cons in there too.
For me the only con is, the waste of ressources (what some would say).
The pro is, I don’t have to maintain another system. No Linux or windows, which needs updates, where I have to assign ressources, etc.
Hi, I just got a new Connbee II, but it disconnects immediately and does not reconnect, the LOG seems fine to me. Any hint?
[s6-init] making user provided files available at /var/run/s6/etc…exited 0.
[s6-init] ensuring user provided files have correct perms…exited 0.
[fix-attrs.d] applying ownership & permissions fixes…
[fix-attrs.d] done.
[cont-init.d] executing container initialization scripts…
[cont-init.d] firmware.sh: executing…
[10:01:36] INFO: GCFFlasher V3_13 © dresden elektronik ingenieurtechnik gmbh
Path | Vendor | Product | Serial | Type
-----------------±-------±--------±-----------±------
/dev/ttyACM0 | 0x1CF1 | 0x0030 | | ConBee II
[cont-init.d] firmware.sh: exited 0.
[cont-init.d] nginx.sh: executing…
[cont-init.d] nginx.sh: exited 0.
[cont-init.d] novnc.sh: executing…
[cont-init.d] novnc.sh: exited 0.
[cont-init.d] done.
[services.d] starting services
[08:01:36] INFO: Running the IKEA OTA updater…
[services.d] done.
[10:01:36] INFO: Websockify waiting for VNC to start
[08:01:36] INFO: Running the deCONZ OTA updater…
[08:01:36] INFO: Starting udevd…
[08:01:36] INFO: Running the OSRAM LEdvance OTA updater…
[10:01:37] INFO: Starting VNC server (local/yes)…
[10:01:37] INFO: Starting websockify…
WebSocket server settings:
- Listen on 127.0.0.1:5901
- Flash security policy server
- Web server. Web root: /usr/share/novnc
- No SSL/TLS support (no cert file)
- proxying from 127.0.0.1:5901 to 127.0.0.1:5900
[10:01:40] INFO: deCONZ waiting for VNC to start
[10:01:41] INFO: Starting the deCONZ gateway…
libpng warning: iCCP: known incorrect sRGB profile
10:01:42:400 HTTP Server listen on address 0.0.0.0, port: 40850, root: /usr/share/deCONZ/webapp/
10:01:42:431 CTRL. 3.27.210:01:42:492 COM: /dev/ttyACM0 / serialno:
10:01:42:492 COM: --dev: /dev/ttyACM0 (ConBee II)
10:01:42:492 ZCLDB init file /data/.local/share/dresden-elektronik/deCONZ/zcldb.txt
[08:01:42] INFO: Starting Nginx…
10:01:42:575 parent process s6-supervise
10:01:42:576 gw run mode: docker/hassio
10:01:42:576 GW sd-card image version file does not exist: /data/.local/share/dresden-elektronik/deCONZ/gw-version
10:01:42:576 DB sqlite version 3.27.2
10:01:42:577 DB PRAGMA page_count: 50
10:01:42:577 DB PRAGMA page_size: 4096
10:01:42:577 DB PRAGMA freelist_count: 3
10:01:42:577 DB file size 204800 bytes, free pages 3
10:01:42:577 DB PRAGMA user_version: 6
2020/08/31 08:01:42 [notice] 325#325: using the “epoll” event method
10:01:42:577 DB cleanup
2020/08/31 08:01:42 [notice] 325#325: nginx/1.14.2
2020/08/31 08:01:42 [notice] 325#325: OS: Linux 5.4.56
2020/08/31 08:01:42 [notice] 325#325: getrlimit(RLIMIT_NOFILE): 1048576:1048576
2020/08/31 08:01:42 [notice] 325#325: start worker processes
2020/08/31 08:01:42 [notice] 325#325: start worker process 1567
10:01:42:577 DB create temporary views
10:01:42:600 don’t close database yet, keep open for 900 seconds
10:01:42:600 started websocket server at port 8081
10:01:42:602 found node plugin: libde_rest_plugin.so - REST API Plugin
10:01:42:604 found node plugin: libde_signal_plugin.so - Signal Monitor Plugin
10:01:45:452 found node plugin: libstd_otau_plugin.so - STD OTAU Plugin
10:01:45:474 COM: /dev/ttyACM0 / serialno:
10:01:45:474 COM: --dev: /dev/ttyACM0 (ConBee II)
10:01:45:629 New websocket 172.30.32.1:50718 (state: 3)
10:01:45:630 New websocket 172.30.32.1:50720 (state: 3)
10:01:47:820 Announced to internet https://phoscon.de/discover
[10:01:52] INFO: Successfully send discovery information to Home Assistant.
10:01:55:540 COM: /dev/ttyACM0 / serialno:
10:01:55:540 COM: --dev: /dev/ttyACM0 (ConBee II)
10:01:56:047 device state timeout ignored in state 2
10:01:57:009 device state timeout ignored in state 2
10:01:57:536 GW update firmware found: /usr/share/deCONZ/firmware/deCONZ_ConBeeII_0x26580700.bin.GCF
10:01:57:536 GW firmware version: 0x00000002
10:01:58:009 device state timeout ignored in state 2
10:01:59:009 device state timeout ignored in state 2
10:02:00:009 device state timeout ignored in state 2
10:02:01:009 device state timeout (handled)
10:02:01:024 device disconnected reason: 1, index: 0
10:02:01:509 wait reconnect 15 seconds
10:02:01:509 void zmMaster::handleStateIdle(zmMaster::MasterEvent) not connected goto OFF state
10:02:01:509 device state timeout ignored in state 1
10:02:55:509 wait reconnect 3 seconds
10:02:56:509 wait reconnect 2 seconds
10:02:57:510 wait reconnect 1 seconds
10:02:57:543 COM: /dev/ttyACM0 / serialno:
10:02:57:544 COM: --dev: /dev/ttyACM0 (ConBee II)
10:02:58:035 device state timeout ignored in state 2
10:02:59:009 device state timeout ignored in state 2
10:02:59:509 reconnect network failed, try later
10:02:59:509 start reconnect to network
10:03:00:009 device state timeout ignored in state 2
10:03:01:008 device state timeout ignored in state 2
10:03:02:013 device state timeout ignored in state 2
10:03:03:008 device state timeout (handled)
10:03:03:030 device disconnected reason: 1, index: 0
10:03:03:511 wait reconnect 15 seconds
10:03:03:513 void zmMaster::handleStateIdle(zmMaster::MasterEvent) not connected goto OFF state
10:03:03:513 device state timeout ignored in state 1
10:03:04:508 failed to reconnect to network try=1
10:03:04:509 wait reconnect 14 seconds
10:03:05:508 wait reconnect 13 seconds
10:03:06:509 wait reconnect 12 seconds
10:03:07:509 wait reconnect 11 seconds
10:03:08:509 wait reconnect 10 seconds
10:03:09:509 failed to reconnect to network try=2
10:03:09:509 wait reconnect 9 seconds
10:03:10:509 wait reconnect 8 seconds
10:03:11:509 wait reconnect 7 seconds
10:03:12:509 wait reconnect 6 seconds
10:03:13:509 wait reconnect 5 seconds
10:03:14:509 failed to reconnect to network try=3
10:03:14:509 wait reconnect 4 seconds
10:03:15:513 wait reconnect 3 seconds
10:03:16:509 wait reconnect 2 seconds
10:03:17:509 wait reconnect 1 seconds
10:03:17:559 COM: /dev/ttyACM0 / serialno:
10:03:17:565 COM: --dev: /dev/ttyACM0 (ConBee II)
10:03:18:059 device state timeout ignored in state 2
10:03:19:009 device state timeout ignored in state 2
10:03:19:509 try to reconnect to network try=4
I am running Home-Assistant on a Synology NAS using Docker. I have a Conbee II Stick and everything worked ok-ish until yesterday. The problem was, that one sensor that was working ok in the phoscon app always said it’s not available in HA, that’s why I updated de deCONZ integration yesterday, and since then it’s not starting any more. It complains about finding the Conbee Stick I guess?!? But that one is still there and everything was working ok-ish until yesterday as I said…
Maybe anybody has any idea? Currently there is only one temperature sensor form Xiaomi and a Button (also Xiaomi) connected to the Conbee stick.
Would be great to get it up and running again 
Here are some logs:
[s6-init] making user provided files available at /var/run/s6/etc...exited 0.
[s6-init] ensuring user provided files have correct perms...exited 0.
[fix-attrs.d] applying ownership & permissions fixes...
[fix-attrs.d] done.
[cont-init.d] executing container initialization scripts...
[cont-init.d] firmware.sh: executing...
[12:54:38] INFO: GCFFlasher V3_13 (c) dresden elektronik ingenieurtechnik gmbh
Path | Vendor | Product | Serial | Type
-----------------+--------+---------+------------+-------
/dev/ttyACM0 | 0x1CF1 | 0x0030 | | ConBee II
[cont-init.d] firmware.sh: exited 0.
[cont-init.d] nginx.sh: executing...
[cont-init.d] nginx.sh: exited 0.
[cont-init.d] novnc.sh: executing...
[cont-init.d] novnc.sh: exited 0.
[cont-init.d] done.
[services.d] starting services
[services.d] done.
[10:54:40] INFO: Running the deCONZ OTA updater...
[10:54:40] INFO: Running the OSRAM LEdvance OTA updater...
[12:54:40] INFO: Waiting for device...
[10:54:40] INFO: Starting udevd...
[10:54:40] INFO: Running the IKEA OTA updater...
[12:54:40] INFO: Websockify waiting for VNC to start
[10:55:40] INFO: Starting Nginx...
[12:55:40] INFO: Starting websockify...
2020/08/31 10:55:40 [notice] 326#326: using the "epoll" event method
2020/08/31 10:55:40 [notice] 326#326: nginx/1.14.2
2020/08/31 10:55:40 [notice] 326#326: OS: Linux 4.4.59+
2020/08/31 10:55:40 [notice] 326#326: getrlimit(RLIMIT_NOFILE): 524288:1048576
2020/08/31 10:55:40 [notice] 326#326: start worker processes
2020/08/31 10:55:40 [notice] 326#326: start worker process 860
WebSocket server settings:
- Listen on 127.0.0.1:5901
- Flash security policy server
- Web server. Web root: /usr/share/novnc
- No SSL/TLS support (no cert file)
- proxying from 127.0.0.1:5901 to 127.0.0.1:5900
172.30.32.2 - - [31/Aug/2020:10:56:27 +0000] "GET /ingress.html HTTP/1.1" 200 2007 "-" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_6) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/13.1.2 Safari/605.1.15"
2020/08/31 10:56:29 [error] 860#860: *1 connect() failed (111: Connection refused) while connecting to upstream, client: 172.30.32.2, server: _, request: "GET /pwa/index.html HTTP/1.1", upstream: "http://127.0.0.1:40850/pwa/index.html", host: "192.168.1.9:8123", referrer: "http://192.168.1.9:8123/api/hassio_ingress/027RHD-Ci2C2zSUbiIClPrdwTN_mJFQ6efuydddzJHc/ingress.html"
172.30.32.2 - - [31/Aug/2020:10:56:29 +0000] "GET /pwa/index.html HTTP/1.1" 502 184 "http://192.168.1.9:8123/api/hassio_ingress/027RHD-Ci2C2zSUbiIClPrdwTN_mJFQ6efuydddzJHc/ingress.html" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_6) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/13.1.2 Safari/605.1.15"
2020/08/31 10:56:31 [error] 860#860: *1 connect() failed (111: Connection refused) while connecting to upstream, client: 172.30.32.2, server: _, request: "GET /pwa/index.html HTTP/1.1", upstream: "http://127.0.0.1:40850/pwa/index.html", host: "192.168.1.9:8123", referrer: "http://192.168.1.9:8123/api/hassio_ingress/027RHD-Ci2C2zSUbiIClPrdwTN_mJFQ6efuydddzJHc/ingress.html"
172.30.32.2 - - [31/Aug/2020:10:56:31 +0000] "GET /pwa/index.html HTTP/1.1" 502 184 "http://192.168.1.9:8123/api/hassio_ingress/027RHD-Ci2C2zSUbiIClPrdwTN_mJFQ6efuydddzJHc/ingress.html" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_6) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/13.1.2 Safari/605.1.15"
[12:57:41] FATAL: No device /dev/serial/by-id/usb-dresden_elektronik_ingenieurtechnik_GmbH_ConBee_II_DE1962646-if00 found!
2020/08/31 10:57:41 [notice] 326#326: signal 15 (SIGTERM) received from 308, exiting
2020/08/31 10:57:41 [notice] 860#860: exiting
2020/08/31 10:57:41 [notice] 860#860: exit
In exit
2020/08/31 10:57:41 [notice] 326#326: signal 17 (SIGCHLD) received from 860
2020/08/31 10:57:41 [notice] 326#326: worker process 860 exited with code 0
2020/08/31 10:57:41 [notice] 326#326: exit
Closing socket listening at 127.0.0.1:5901
[cont-finish.d] executing container finish scripts...
[cont-finish.d] done.
[s6-finish] waiting for services.
[s6-finish] sending all processes the TERM signal.
[s6-finish] sending all processes the KILL signal and exiting.
Hi, did you deactivate vnc? if not hen try that. There are some posts regarding this above.
This bug is the reason I moved to bare metal.
it has never been resolved that randomly the by-id device link does not work in the Addon
You only need to use by-id link if you have multiple devices that race with each other during boot to become /dev/ttyACM0 and /dev/ttyACM1.
If you only have one USB device called /dev/ttyACM0 then use that in the Addon config and this problem goes away
YES!!! @KennethLavrsen THANK you very much! That did the trick! Interessing though… cause my Synology NAS only has the Conbee Stick plugged in, nothing else. Anyway… now it starts again! Will have to setup everything again, bug hey
Cool! Thanks!
yes, that works in my config too, much better and easier than the long explicit line…
since I am new to the Deconz and add-on, please forgive me my noob questions:
on the Ikea motion sensor, the ‘detected’ state is permanent, while I can see the motion being detected in the log. It simple never returns to a not detected state. Is this as per design:
secondly, this battery sensor doesnt have a state at all ?
and
lastly for now: as you can see the bottom sensor doesnt have a state, but, when clicking this Ikea remote switch, very briefly and faintly the correct action is shown. Shouldnt this last action remain in the state until the next has occurred?
thanks for having a look
Motion sensors report on and off. They are binary sensors.
Remote controls (switches with buttons) are stateless. They create events that can be viewed as deconz_event. There are 100s of questions just in this thread on this.
Push buttons do not have a state. I know it is tempting to see them as ON as long as you hold the button pressed and OFF when you let go of the button. But most Zigbee remotes do not work like this. They send events when you press and when you release. They also send events for long press and some even have events for double press. So they are not binary. They often have 3 or 4 different event types. That is why these things have to be implemented as event driven devices and not devices with a binary state.
You could then say that maybe it should be a sensor with multiple values. But they does not work either in all cases because the events are not always mutually excluded
All the motion sensors still have sensors defined for battery.
I have seen issues with pairing motion sensors with Deconz lately. They pair but often the other sensors are not registered. I have seen it lately with the Philips remote.
This is not an Addon related issue. This is something in Deconz itself. I normally resolve it by keeping on adding the device until I see all the sensors working. Sometimes it repairs itself it seems if you wait. However I have seen this not being the case
See my bug report
Thanks for your reply! much appreciated.
sure, I understand that. Have a plentitude of them in the system ;-), was only now experimenting with a new Conbee II stick. Thats why I dont understand why it keeps showing ‘detected’ for ages, and doesnt change to clear after a few seconds.
Yep, forgot about that, still used to the former Hue CC sensors for the switches… Still, the very faint display of the actual action could be clearer.
Guess the CC event sensor will provide some assistance.
Will try to reinstall the motion sensor and see what happens. Any tips for deleting a sensor in the best way?
btw, Ive added to your issue, maybe the stale bot wont declare Stale so quickly 