Z-Wave issues since some months including E5 for LC-13

I am experiencing notorious Z-Wave issues with battery devices since the heating period started this year. My Danfoss LC13 thermostats (which worked just fine for 5 years until now) started become unresponsive and the same goes for the battery wall switches I have. They work fine for some days or even weeks eventually stop working until they are power cycled. The LC 13 devices show the dreaded E5-Error. I have searched the forums for this issue but could not really find a solution for this. I have ex- and included, hard and soft reset, deleted and re-initialized the devices, but nothing helped for long. The only thing that’s different now is that I have ghost devices in the controller which can’t be removed because the exclusion did not work or I did a factory reset too early.

Half of my non-battery devices also show Status: OFFLINE - COMMUNICATION_ERROR Node is not communicating with controller in Paper-UI but they are still accepting commands.

Two days ago suddenly the whole z-wave Network become unresponsive and I had to restartet OH to make it functional again. The Logfile contained this exception:

2020-11-27 08:53:03.144 [WARN ] [WaveSerialHandler$ZWaveReceiveThread] - Exception during ZWave thread.
java.lang.IllegalStateException: Queue full
    at java.util.AbstractQueue.add(AbstractQueue.java:98) ~[?:1.8.0_201]
    at java.util.concurrent.ArrayBlockingQueue.add(ArrayBlockingQueue.java:312) ~[?:1.8.0_201]
    at org.openhab.binding.zwave.internal.protocol.ZWaveTransactionManager.processReceiveMessage(ZWaveTransactionManager.java:447) ~[bundleFile:?]
    at org.openhab.binding.zwave.internal.protocol.ZWaveController.incomingPacket(ZWaveController.java:1078) ~[bundleFile:?]
    at org.openhab.binding.zwave.handler.ZWaveControllerHandler.incomingMessage(ZWaveControllerHandler.java:444) ~[bundleFile:?]
    at org.openhab.binding.zwave.handler.ZWaveSerialHandler$ZWaveReceiveThread.run(ZWaveSerialHandler.java:454) [bundleFile:?]

I have the feeling my z-wave network is slowly falling apart.

I am using a VIS_ZU1401 stick and the latest OH 2.5 release from the Debian repository.

Is there anything I can do?

I recommend to get rid of the ghost nodes in a first step. You may find information about the process in this thread .


In addition to remobving the ghost nodes, I have only seen queue full messages when the daily heal is enabled. If you still have slowness after removing the ghosts, disable the daily heal in the controller settings and restart OH. After that, if you change anything in the network, like powering a device off, moving a device, or replacing a device, manually heal your mains powered devices.

1 Like

Thanks for the help, I will try that. I wonder what has changed, though, as everything worked just fine for several years and my devices remained mostly the same in this time.

Controllers or a device could have gone bad. Power cycle every device. Shut down the equipment running OH and then apply power… some USB ports do not power cycle during a restart. RF is very finicky, so moving anything in the house could cause trouble. It is very hard to say what has happened without seeing debug logs from the binding or without setting up a sniffer.

Thanks for the help! Disabling the auto heal fixed my Thermostats. No E5 since 10 days or so. I remember @chris writing somewhere else about some unnecessary feature during auto heal which causes problems and which he wants to remove for OH 3.0, so maybe I try enabling auto heal again after the update.

I did not exclude the ghosts yet, though, and woke up today with the z-wave queue being full again. So the two issues I noticed do not seem to be related. That’s the next thing I’ll do.

1 Like