KNX - Error in logs

Third party integrations
1

Hi,

I am troubleshooting why my KNX connection is falling out every day. It suddently stopes working and every thing works again after a reboot of HA.

I have quite a lot of errors in the log with my thermostas. I have disabled most of them, to make troubleshooting easier. So now only one is active.

Here is the error:

Logger: xknx.log
Source: /usr/local/lib/python3.8/site-packages/xknx/remote_value/remote_value.py:190
First occurred: 2:07:59 PM (5 occurrences)
Last logged: 3:25:05 PM

  • Could not sync group address ‘4/2/3’ (Termostat Hall 2.ETG - Current Temperature)
  • Could not sync group address ‘0/2/1’ (Termostat Hall 2.ETG Mode - Operation mode)
  • Could not sync group address ‘6/2/3’ (Termostat Hall 2.ETG - Target temperature)

Is this normal, everything works fine for about 6-20 hours. I can read temperature, change operation mode and set new target? But I get the errors several times an hour and the communcation on the knx bus stoppes after some hours.

Any ideas?

Best regards Jon

1 Like
2

Hi!
No this is not normal. First of all, which HA version are you running?
So after 6 hours you can’t switch devices anymore? Or is it just the errors, but communication still works? If it doesn’t work, I assume there are more errors in the logs.
Are there more connections to your interface than HA (OpenHAB, ETS etc. ) and if yes, can your interface handle it?

This is a great idea! Can you post it’s yaml configuration and ETS screenshots of the group addresses / communication objects involved (with flags column)?

3

Hi,

thanks for your reply!

I am running this version:

Hassio
Hassio3102×1201 146 KB

So after 6 hours you can’t switch devices anymore?
Answer: After a period of time, between 6-20 hours the communication with the KNX interface does not work at all, can control lights, switches and thermostats. I can get it working again when restarting the Server management (Configuration - Server Controls - Server management). But I can connect to my KNX system from ETS.

Or is it just the errors, but communication still works? If it doesn’t work, I assume there are more errors in the logs.
Answer: Everything is down(KNX). There is only errors on the thermostats, have not found any thing else. I tried to disconnect the RJ45 and connect it again on the interface, HA reported then in the logs that the KNX heart beat was lost, and reconnects to it. So this is a also a way to make it work again.

Are there more connections to your interface than HA (OpenHAB, ETS etc. ) and if yes, can your interface handle it?
Answer: Its a Gira 2168 00, it can handle 4 devices at the same time(tunneling).

I have tried with just knx in the config and also static, tunneling info. Same result.

See picture in new post.

Climate setup:

See picture in new post.

Regarding the thermostat, it is strange that it reports errors when it actually works, it seams that it happens to some specific time every hour. Se screenshot under:

See picture in new post.

I am very happy to get this solved, this drives me nuts!
Best regards Jon

4

KNX yaml

5

Climate
Climate938×601 65.3 KB

6

Errors
Errors2239×1503 724 KB

7

yes, every 60 minutes after the last incoming telegram.
6/2/3 , 4/2/3 and 0/2/1 are readable from ETS Group Monitor?

This is what I was looking for. It only occurs when you unplug the cable? Not on your “normal” knx connection losses?

pychromecast also has heartbeat timeout. So I’m guessing this is a network problem.

PS: Please don’t post screenshots of text - especially logs. Copy and paste it - it’s much better readable then.

8

Hi,

Thanks.

The heart beat message only appears in the logs when I disconnect the RJ45 and reconnect it, not when I loose the connection over time.

The telegrams is readable from Group monitor.

I can control KNX from HA and simultaneously enable Group monitor in ETS, but it is not possible to activate BUS monitor when HA is connected to KNX interface. If I reboot the interface and activate BUS monitor in ETS, that works, but then HA can not connect? Is this correct behavior?

Any ideas what could cause this problems in the network?

I am trying a new setup now, with static ip on the interface and HA(written in the yaml knx setup).

Really appriciate your input!

Regards Jon

9

Yes this is normal. You most likely only need the Group Monitor.

It’s worth a shot but I wouldn’t bet it will change anything.

Not at all. Since you said you unplug the cable I guess there is no WiFi involved…

You can try to set up debug log level for the "xknx.log" logger and wait for a connection loss. And when you don't find anything suspicious "xknx.knx" too, but this is a lot of information to parse...

In the coming release 2021.2 (core) there are some changes to knx reconnection logic. Maybe this will help you.
Without any error messages we can’t really locate the actual problem :thinking: (and I don’t think the “could not sync” errors are related.)