Network Error with KNX Integration

[LukasGrebe]

The problem

I've noticed the state of KNX entities sometimes fails to sync. This is noticeable in the History chars by tiny gray bars indicating "unavailable"

From everything I've gathered this appears to be a network issue between HomeAssistant and the KNX IP Interface. All components are connected via ethernet so this doesn't appear to be a wifi connectivity issue.
The issue appears randomly throughout the day

What version of Home Assistant Core has the issue?

2022.4.6

What was the last working version of Home Assistant Core?

No response

What type of installation are you running?

Home Assistant OS

Integration causing the issue

knx

Link to integration documentation on our website

https://www.home-assistant.io/integrations/knx/

Diagnostics information

config_entry-knx-c239f84ea9bd9e7c1b40a499ab89f87c.json.txt

Example YAML snippet

No response

Anything in the logs that might be useful for us?

Logger: xknx.log
Source: /usr/local/lib/python3.9/site-packages/xknx/remote_value/remote_value.py:254 
First occurred: 13. Mai 2022, 14:43:14 (3397 occurrences) 
Last logged: 07:24:19

Could not sync group address '6/1/12' (Hz Ventil aaa - Value)
Could not sync group address '6/1/14' (Hz Ventil bbb - Value)
Could not sync group address '6/1/17' (Hz Ventil ccc - Value)
Could not sync group address '6/1/21' (Hz Ventil ddd - Value)
Could not sync group address '30/7/4' (Diagnose Text - Value)

Logger: xknx.log
Source: /usr/local/lib/python3.9/site-packages/xknx/core/value_reader.py:57 
First occurred: 13. Mai 2022, 14:43:14 (3397 occurrences) 
Last logged: 07:24:19

Error: KNX bus did not respond in time (2.0 secs) to GroupValueRead request for: 6/1/12
Error: KNX bus did not respond in time (2.0 secs) to GroupValueRead request for: 6/1/14
Error: KNX bus did not respond in time (2.0 secs) to GroupValueRead request for: 6/1/17
Error: KNX bus did not respond in time (2.0 secs) to GroupValueRead request for: 6/1/21
Error: KNX bus did not respond in time (2.0 secs) to GroupValueRead request for: 30/7/4


Logger: xknx.log
Source: /usr/local/lib/python3.9/site-packages/xknx/io/tunnel.py:359 
First occurred: 13. Mai 2022, 14:43:07 (145 occurrences) 
Last logged: 07:03:51

Received DisconnectRequest from tunnelling sever.


Logger: xknx.log
Source: /usr/local/lib/python3.9/site-packages/xknx/io/tunnel.py:306 
First occurred: 13. Mai 2022, 17:12:13 (688 occurrences) 
Last logged: 07:22:37

L_DATA_CON Data Link Layer confirmation timed out for <Telegram direction="Outgoing" source_address="1.1.31" destination_address="8/2/72" payload="<GroupValueRead />" />
L_DATA_CON Data Link Layer confirmation timed out for <Telegram direction="Outgoing" source_address="1.1.31" destination_address="8/2/73" payload="<GroupValueRead />" />
L_DATA_CON Data Link Layer confirmation timed out for <Telegram direction="Outgoing" source_address="1.1.31" destination_address="8/2/74" payload="<GroupValueRead />" />
L_DATA_CON Data Link Layer confirmation timed out for <Telegram direction="Outgoing" source_address="1.1.31" destination_address="1/2/14" payload="<GroupValueRead />" />
L_DATA_CON Data Link Layer confirmation timed out for <Telegram direction="Outgoing" source_address="1.1.31" destination_address="1/2/12" payload="<GroupValueRead />" />


Logger: async_upnp_client.profiles.profile
Source: /usr/local/lib/python3.9/site-packages/async_upnp_client/profiles/profile.py:272 
First occurred: 14. Mai 2022, 16:05:32 (1 occurrences) 
Last logged: 14. Mai 2022, 16:05:32

Failed (re-)subscribing to: uuid:c75d7ae0-8759-4d65-b69e-0cc0133ae401, reason: UpnpConnectionTimeoutError('', None)


Logger: xknx.log
Source: /usr/local/lib/python3.9/site-packages/xknx/io/tunnel.py:395 
First occurred: 14. Mai 2022, 01:04:44 (9 occurrences) 
Last logged: 14. Mai 2022, 14:20:25

Heartbeat to KNX bus failed. No answer from tunneling server.

Additional information

I've truncated the KNX Config in the diagnostics info for privacy

[probot-home-assistant]

knx documentation
knx source
_{^{(message by IssueLinks)}}

[probot-home-assistant]

Hey there @Julius2342, @farmio, @marvin-w, mind taking a look at this issue as it has been labeled with an integration (knx) you are listed as a code owner for? Thanks!
_{^{(message by CodeOwnersMention)}}

[farmio]

Hi 👋!
This seems like a duplicate of #59170 except we think we have solved this in 2022.4 🤔

Please upgrade HA to the latest version and see if it persists.

What Knx Interface are you using?

[LukasGrebe]

Hi! Thanks so much. Ive updated HomeAssistant Core to 2022.5.4 and haven't yet seen the effect but will gather some more observations.

Im also still seeing some "did not respond in time" and "failed to sync" but i think these might be configuration errors of a few knx adresses.

Ill have another look and close the ticket thereafter.

[LukasGrebe]

unfortunately i'm still seeing the effect after updating on Sunday:

[farmio]

Oh no!
Unfortunately the source of such issues probably isn't easy to find.
For a start, can you get me a whole (untruncated) HA log file with some debug information:

# configuration.yaml
logger:
  default: info
  logs:
    xknx.log: debug
    xknx.knx: debug

You can reach me on discord via farmio#5918 on HA server or xknx server if you are concerned about publicly uploading it.

[fuchS31337]

Same here :/
Problem started to appear around 6th to 8th of may with 22.5.0 or 5.1? persists until 5.4
Until I lowered the telegramrate my Weizierl gateway requested a tunnel reset (according to HA log). That happens nearly perfect every hour. After I lowered the rate to 5, only the grey lost connections appear. Telegramrate spikes to 60 percent and the ETS Monitor wents wild. Never happened before 6th of may.

Gone back to 5.1 (my oldest backup) no change.
Tried to disable the broken Govee and DWD integrations, no change.

Hope this can be resolved, its scary when your blinds start moving in the middle of the night ;-)

[farmio]

We would need much more information to resolve this. Like a Wireshark packet capture or ETS log and detailed HA logs.

Without more information we can't possibly know what is happening here.

Lowering the telegram rate wont really solve a problem.

[fuchS31337]

I started logging both