All devices not available for 12 seconds

[marcel51090]

The problem

Since the update to 2024.6 and also 2024.6.1 and 2, I have the problem that my zigbee devices that run over Zha are constantly unavailable for 12 seconds.

What version of Home Assistant Core has the issue?

2024.6.2

What was the last working version of Home Assistant Core?

2024.5.2

What type of installation are you running?

Home Assistant OS

Integration causing the issue

ZHA

Link to integration documentation on our website

No response

Diagnostics information

No response

Example YAML snippet

No response

Anything in the logs that might be useful for us?

logger:%20bellows.ezsp%0AQuelle%3A%20/usr/local/lib/python3.12/site-packages/bellows/ezsp/__init__.py:496%0AErstmals%20aufgetreten:%2009:14:22%20(1%20Vorkommnisse)%0AZuletzt%20protokolliert:%2009:14:22%0A%0AException%20running%20handler%0ATraceback%20(most%20recent%20call%20last):%0A%20%20File%20%22/usr/local/lib/python3.12/site-packages/bellows/ezsp/__init__.py%22,%20line%20494,%20in%20handle_callback%0A%20%20%20%20handler(*args)%0A%20%20File%20%22/usr/local/lib/python3.12/site-packages/bellows/zigbee/application.py%22,%20line%20626,%20in%20ezsp_callback_handler%0A%20%20%20%20self.connection_lost(args%5B0%5D)%0A%20%20File%20%22/usr/local/lib/python3.12/site-packages/zigpy/application.py%22,%20line%20688,%20in%20connection_lost%0A%20%20%20%20self.listener_event(%22connection_lost%22,%20exc)%0A%20%20File%20%22/usr/local/lib/python3.12/site-packages/zigpy/util.py%22,%20line%2050,%20in%20listener_event%0A%20%20%20%20for%20listener,%20include_context%20in%20self._listeners.values():%0ARuntimeError:%20dictionary%20changed%20size%20during%20iteration%0A

Additional information

No response

[home-assistant]

Hey there @dmulcahey, @Adminiuga, @puddly, @TheJulianJES, mind taking a look at this issue as it has been labeled with an integration (zha) you are listed as a code owner for? Thanks!

Code owner commands

Code owners of zha can trigger bot actions by commenting:

• @home-assistant close Closes the issue.
• @home-assistant rename Awesome new title Renames the issue.
• @home-assistant reopen Reopen the issue.
• @home-assistant unassign zha Removes the current integration label and assignees on the issue, add the integration domain after the command.
• @home-assistant add-label needs-more-information Add a label (needs-more-information, problem in dependency, problem in custom component) to the issue.
• @home-assistant remove-label needs-more-information Remove a label (needs-more-information, problem in dependency, problem in custom component) on the issue.

_{^{(message by CodeOwnersMention)}}

---

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

[puddly]

Can you enable debug logging for the ZHA integration, wait for your devices to stop working, and disable debug logging to download the log? Without the log it's difficult to say what isn't working.

[marcel51090]

No problem. The problem should be present in the log file (10:39 was the problem)

home-assistant.log

[madspeed]

I can confirm the issue, however in my case the period is different from 10 to 12 seconds.
The case is definitely with update 2024.6.1 and 2024.6.2. Tried to restore backup 2024.5.5 and everything is working perfectly, after upgrading to 2024.6.2 the problem appeared again. Zigbee coordinator is Skyconnect.

[puddly]

@marcel51090 I'm seeing no mention of zha in your debug log. You can enable ZHA debug logging like this:

@madspeed Can you also create a debug log that includes one of the resets?

[madspeed]

Hope it is there
home-assistant_2024-06-12T20-23-59.424Z.log

[puddly]

@madspeed Strange. It looks like your firmware is crashing after a device re-joins the network multiple times, very quickly:

2024-06-12 23:20:20.831 DEBUG (MainThread) [zigpy.application] Connection to the radio has been lost: <NcpResetCode.RESET_ASSERT: 6>

Can you try updating to a newer firmware? I have a build of 7.4.2.0 that you can flash: NabuCasa/silabs-firmware-builder#71 (comment)

[EmilyNerdGirl]

I've been seeing this as well, going to try out the new firmware.

Edit: Done

[madspeed]

@madspeed Strange. It looks like your firmware is crashing after a device re-joins the network multiple times, very quickly:

2024-06-12 23:20:20.831 DEBUG (MainThread) [zigpy.application] Connection to the radio has been lost: <NcpResetCode.RESET_ASSERT: 6>

Can you try updating to a newer firmware? I have a build of 7.4.2.0 that you can flash: NabuCasa/silabs-firmware-builder#71 (comment)

Unfortunately upgrade to 7.4.2.0 build 0 did not solve the problem.

[cobirnm]

I have the same exact problem but with an exception. If I activate the debug mode for ZHA the problem does not occur. When I turn off debug I start having all ZHA integration unavailable every 3 or 4 minutes. On 2024.5.5 this does not happen.

[puddly]

@cobirnm Can you leave debug mode on until the problem occurs again?

[cobirnm]

@cobirnm Can you leave debug mode on until the problem occurs again?

Please see the log.
Never the less if I keep ZHA on debug mode the error does not happen. I had to put debug mode only for zigpy

Log zigpy.txt

[puddly]

Zigpy debug logging won't be of use here, bellows is the library responsible for communicating with your stick.

I've been able to reproduce this problem and will update this issue once it's fixed.

[CordoWEB]

I noticed the same problem on 2024.6.2 version.
I'm going back to 2024.5.5 version to see what happens.