Docker 2.4 to 2.5 zwave stick failing

Hi!

I have the aeotec gen5 stick and after upgrade from 2.4 to 2.5 the state of the stick changes very frequently from online to offline. and I dont see an obvious reason to why.
I have tried rebooting the container and also powercycled the whole docker machine, disconnecting/reconnecting the stick. but Nothing helps and I am out of ideas…

The upgrade:
As I am not very sofisticated yet i.e using docker compose. I use a oneliner. and simply stopped the old container pruned it and changed “openhab/openhab:2.4.0” into “openhab/openhab:2.5.0”

snippet from the eventlog

2019-12-28 18:05:40.858 [me.event.ThingUpdatedEvent] - Thing 'zwave:serial_zstick:689f3d91' has been updated.
2019-12-28 18:05:40.860 [hingStatusInfoChangedEvent] - 'zwave:device:689f3d91:node8' changed from OFFLINE (BRIDGE_OFFLINE): Controller is offline to ONLINE
2019-12-28 18:05:40.913 [hingStatusInfoChangedEvent] - 'zwave:device:689f3d91:node9' changed from ONLINE to ONLINE: Node initialising: PING
2019-12-28 18:05:40.937 [hingStatusInfoChangedEvent] - 'zwave:device:689f3d91:node8' changed from ONLINE to ONLINE: Node initialising: PING
2019-12-28 18:05:40.979 [hingStatusInfoChangedEvent] - 'zwave:device:689f3d91:node6' changed from ONLINE to ONLINE: Node initialising: PING
2019-12-28 18:05:41.040 [hingStatusInfoChangedEvent] - 'zwave:device:689f3d91:node19' changed from OFFLINE (BRIDGE_OFFLINE): Controller is offline to ONLINE
2019-12-28 18:05:41.062 [hingStatusInfoChangedEvent] - 'zwave:device:689f3d91:node17' changed from OFFLINE (BRIDGE_OFFLINE): Controller is offline to ONLINE
2019-12-28 18:05:41.084 [hingStatusInfoChangedEvent] - 'zwave:device:689f3d91:node16' changed from OFFLINE (BRIDGE_OFFLINE): Controller is offline to ONLINE
2019-12-28 18:05:41.130 [hingStatusInfoChangedEvent] - 'zwave:device:689f3d91:node14' changed from OFFLINE (BRIDGE_OFFLINE): Controller is offline to ONLINE
2019-12-28 18:05:41.136 [hingStatusInfoChangedEvent] - 'zwave:device:689f3d91:node12' changed from OFFLINE (BRIDGE_OFFLINE): Controller is offline to ONLINE
2019-12-28 18:05:41.141 [hingStatusInfoChangedEvent] - 'zwave:device:689f3d91:node13' changed from OFFLINE (BRIDGE_OFFLINE): Controller is offline to ONLINE
2019-12-28 18:05:41.144 [hingStatusInfoChangedEvent] - 'zwave:device:689f3d91:node10' changed from OFFLINE (BRIDGE_OFFLINE): Controller is offline to ONLINE
2019-12-28 18:05:41.148 [hingStatusInfoChangedEvent] - 'zwave:device:689f3d91:node18' changed from OFFLINE (BRIDGE_OFFLINE): Controller is offline to ONLINE
2019-12-28 18:05:41.150 [hingStatusInfoChangedEvent] - 'zwave:device:689f3d91:node15' changed from OFFLINE (BRIDGE_OFFLINE): Controller is offline to ONLINE
2019-12-28 18:05:41.152 [hingStatusInfoChangedEvent] - 'zwave:device:689f3d91:node14' changed from ONLINE to ONLINE: Node initialising: PING
2019-12-28 18:05:41.156 [hingStatusInfoChangedEvent] - 'zwave:device:689f3d91:node11' changed from OFFLINE (BRIDGE_OFFLINE): Controller is offline to ONLINE
2019-12-28 18:05:41.163 [hingStatusInfoChangedEvent] - 'zwave:device:689f3d91:node15' changed from ONLINE to ONLINE: Node initialising: PING
j

-J

Check addons.cfg and .addons.config for invalid addons. Restdocs moved from misc to ui and yahooweather is no longer valid. The loader reloads all addons every minute trying to load the invalid one.

is this the correct place:
“/home/openhab/conf/services” the addons file there has no addons. I do have the complaint about the rest-doc but never thought those could impact each other :confused:
Went to the paper gui and addons and installed the rest through there(in the new location) Where do I remove the reference to the old if not in those files?

Timewise I dont see exact relation but I am open for Anything at this point

2019-12-28 18:44:48.128 [ERROR] [core.karaf.internal.FeatureInstaller] - Failed installing 'openhab-misc-restdocs'
2019-12-28 18:45:55.386 [INFO ] [ing.zwave.handler.ZWaveSerialHandler] - Stopped ZWave serial handler

Not sure about Docker but openhabian is:

Check addons.cfg at etc/openhab2/services/addons.cfg

and addons.config at

/var/lib/openhab2/config/org/openhab/addons.config

If you change addons.config make sure to stop OH first sudo systemctl stop openhab2

Link to help with docker file location Using Sonoff Power Switches with Tasmota firmware and openHAB2 MQTT2 binding

Thankyou for the reminder to stop it before. I stopped the container and edited the file and now its “booting” let’s see what happens.

Would be so sweet if the zwave stick behaves after this…

Can you post the contents of your addons.config or let me know what the first line in the file is please? Really just curious about the revision:=L"somenumber" part.

Sure!

:org.apache.felix.configadmin.revision:=L"1"
binding="zwave,network,onkyo,unifi,astro,chromecast"
...

After the reboot is complete and you still have issues I would change that 1 to a 2 (stop OH first,:laughing:) clean cache, reboot and see what happens.

On openhabian I have L"35"

Make sense when you say it… changing the revision… but then you lost me :smiley: clear cache. ehm not sure how to do that. Though! fingers crossed… The stick has been alive almost 8 minutes now without any hickup. I saw another err in the openhab.log though relating to the new rule engine. But I am assuming something is configured wrong and or missing from my part.

2019-12-28 19:00:39.816 [ERROR] [org.openhab.core.ephemeris          ] - bundle org.openhab.core.ephemeris:2.5.0 (143)[org.openhab.ephemeris(58)] : The activate method has thrown an exception
java.lang.IllegalArgumentException: No enum constant java.time.DayOfWeek.SATURDAY,SUNDAY

Stop OH and use sudo openhab-cli clean-cache

:+1:

That warning is nothing to worry about and will go away after you set your location info in PaperUI and restart OH a few times. Location really has nothing to do with it but I recomend setting it either way.

Oki Thx!
Thinking of maybe changing that revision. It doesnt complain anymore but it seems that something is still abit broken ish… I cant control any of the lights but I get no error… so I’ll try that revision change. still not sure how to apply the cache clean in the container

Did you try openhab-cli clean-cache ?

Wait I found this, great author as well.:rofl:

1 Like

I am not starting the container with bash possibility so I am not sure how to enter it to be honest.

haha! :sweat_smile: ok, then its more clear!.. I will change rev and clean the cache

1 Like

uhm, I dont really have a file to clear, there are a bunch of directories though under each of these dir’s.

/home/openhab/userdata/cache$ ls -lah
total 16K
drwxr-xr-x   4 openhab openhab 4.0K Dec 28 15:59 .
drwxr-xr-x  11 openhab openhab 4.0K Dec 28 15:56 ..
drwxr-xr-x 273 openhab openhab 4.0K Dec 28 18:59 org.eclipse.osgi
drwxr-xr-x   2 openhab openhab 4.0K Dec 28 15:59 org.openhab.voice.googletts

I have org.eclipse.osgi and a log in my cache. Here is the old way to clean cache and tmp (btw openhab-cli does both)

sudo rm -rf /var/lib/openhab2/cache/* && sudo rm -rf /var/lib/openhab2/tmp/*

1 Like

hmm, so you do delete those. and I guess they are re-created after(?)
sounds abit scary, eventhough they are in cache and tmp directories

That are recreated when starting openHAB. The initial startup will take longer because of this.

2 Likes

Its safe to run and as @Bruce_Osborne mentioned above startup will be a bit slower.

Thanks for jumping in Bruce…Hard to keep up with trying to help 4 others at once.:crazy_face:

2 Likes

thankful for all the help!!

I deleted the files under that directory and started the container again. but I still cannot control anything via zwave. Strangely it seem to receive notifications from my temperature sensor so it’s not completely dead… and the zwave stick is not constantly restarting anymore :slight_smile: so its a step in the right direction