ZHA network, devices becomes uncontrollable

[Esp32-zapper]

The problem

I NEED HELP!

My Zigbee network has run like clockwork for over a year but the last couple of months strange errors has come creeping in and after 2025.4.0 it is now PANIC! ZHA works very randomly. Things become uncontrollable (not unreachable). Been working like crazy last night, after the update, to get some balance again. Been re-pairing devices that become uncontrollable and that works.... FOR A WHILE. After a restart a whole new set of devices has become uncontrollable and re-pairing them solves the problem until you restart and then a random set of devices becomes uncontrollable again and could be one that I just re-paired. I can't se any pattern of brands that are more or less likely to have problems. I get problem with Aqara thermostats Tuya light strips Sonoff motion sensor Aqara in-wall light switches (no neutral) Lidl Smartplugs Lidl Smart buttons Among others

This morning an update for my SkyConnect became available and I thought that might sove the issues, so I did the update and restarted. Did absolutely nothing. Same amount of problems and same pattern.

A clue can also be that sometimes a problematic device, after the error message comes up some 30 seconds after the action, starts working and works flawlessly when you turn on/off but after a few minutes that device is uncontrollable again.

Naturally I have searched for colliding wifi networks and the SkyConnect is on an extension cable, well away from anything that can interfer. No issues there and my WiFis are set to fixed channels in the router.

I NEED HELP URGENTLY! (Wifie is really pissed)

What version of Home Assistant Core has the issue?

core-2025.4.0

What was the last working version of Home Assistant Core?

core-2025.1 (creeping problem so it's an estimate)

What type of installation are you running?

Home Assistant OS

Integration causing the issue

ZHA

Link to integration documentation on our website

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

Diagnostics information

Loggare: homeassistant.components.websocket_api.http.connection Källa: components/websocket_api/commands.py:245 integration: Home Assistant WebSocket API (dokumentation, ärenden) Inträffade först: 07:22:07 (3 tillfällen) Senast loggade: 07:22:45

[140384435382320] Unexpected exception Traceback (most recent call last): File "/usr/local/lib/python3.13/site-packages/zha/zigbee/cluster_handlers/init.py", line 68, in wrap_zigpy_exceptions yield File "/usr/local/lib/python3.13/site-packages/zha/zigbee/cluster_handlers/init.py", line 86, in wrapper return await RETRYABLE_REQUEST_DECORATOR(func)(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.13/site-packages/zigpy/util.py", line 136, in retry return await func() ^^^^^^^^^^^^ File "/usr/local/lib/python3.13/site-packages/zigpy/zcl/init.py", line 378, in request return await self._endpoint.request( ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ...<9 lines>... ) ^ File "/usr/local/lib/python3.13/site-packages/zigpy/endpoint.py", line 270, in request return await self.device.request( ^^^^^^^^^^^^^^^^^^^^^^^^^^ ...<11 lines>... ) ^ File "/usr/local/lib/python3.13/site-packages/zigpy/device.py", line 378, in request await send_request() File "/usr/local/lib/python3.13/site-packages/zigpy/application.py", line 841, in request await self.send_packet( ...<14 lines>... ) File "/usr/local/lib/python3.13/site-packages/bellows/zigbee/application.py", line 929, in send_packet raise zigpy.exceptions.DeliveryError( f"Failed to deliver message: {send_status!r}", send_status ) zigpy.exceptions.DeliveryError: Failed to deliver message: <sl_Status.ZIGBEE_DELIVERY_FAILED: 3074>

The above exception was the direct cause of the following exception:

Traceback (most recent call last): File "/usr/src/homeassistant/homeassistant/components/zha/helpers.py", line 1352, in handler return await func(self, *args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/src/homeassistant/homeassistant/components/zha/light.py", line 184, in async_turn_on await self.entity_data.entity.async_turn_on( ...<6 lines>... ) File "/usr/local/lib/python3.13/site-packages/zha/application/platforms/light/init.py", line 399, in async_turn_on result = await self._on_off_cluster_handler.on() ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.13/site-packages/zha/zigbee/cluster_handlers/init.py", line 85, in wrapper with wrap_zigpy_exceptions(): ~~~~~~~~~~~~~~~~~~~~~^^ File "/usr/local/lib/python3.13/contextlib.py", line 162, in exit self.gen.throw(value) ~~~~~~~~~~~~~~^^^^^^^ File "/usr/local/lib/python3.13/site-packages/zha/zigbee/cluster_handlers/init.py", line 77, in wrap_zigpy_exceptions raise ZHAException(message) from exc zha.exceptions.ZHAException: Failed to send request: Failed to deliver message: <sl_Status.ZIGBEE_DELIVERY_FAILED: 3074>

The above exception was the direct cause of the following exception:

Traceback (most recent call last): File "/usr/src/homeassistant/homeassistant/components/websocket_api/commands.py", line 245, in handle_call_service response = await hass.services.async_call( ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ...<7 lines>... ) ^ File "/usr/src/homeassistant/homeassistant/core.py", line 2794, in async_call response_data = await coro ^^^^^^^^^^ File "/usr/src/homeassistant/homeassistant/core.py", line 2837, in _execute_service return await target(service_call) ^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/src/homeassistant/homeassistant/helpers/service.py", line 1007, in entity_service_call single_response = await _handle_entity_call( ^^^^^^^^^^^^^^^^^^^^^^^^^^ hass, entity, func, data, call.context ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ) ^ File "/usr/src/homeassistant/homeassistant/helpers/service.py", line 1079, in _handle_entity_call result = await task ^^^^^^^^^^ File "/usr/src/homeassistant/homeassistant/components/light/init.py", line 648, in async_handle_light_on_service await light.async_turn_on(**filter_turn_on_params(light, params)) File "/usr/src/homeassistant/homeassistant/components/zha/helpers.py", line 1354, in handler raise HomeAssistantError(err) from err homeassistant.exceptions.HomeAssistantError: Failed to send request: Failed to deliver message: <sl_Status.ZIGBEE_DELIVERY_FAILED: 3074>

Example YAML snippet

Anything in the logs that might be useful for us?

Additional information

My zha settings in configuration.yaml

ZHA config zha: custom_quirks_path: /config/custom_zha_quirks/ zigpy_config: ota:

enabled: true

  extra_providers:
    - type: zigpy_remote
      url: https://raw.githubusercontent.com/Koenkk/zigbee-OTA/master/index.json
    - type: z2m

path: /media/Media/HA/ZHA-update

otau_directory: /media/Media/HA/ZHA-update

source_routing: false
ezsp_config:
  CONFIG_MAX_END_DEVICE_CHILDREN: 0

[home-assistant[bot]]

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)}}

[Esp32-zapper]

I think this might be related but still clueless. Have about hundred devices, so very hard to know witch one it refers to. Wound be much more user friendly to actually give the name of the device in the message instead of only a hex address. Is there a way to search for a device by Zigbee hex address? Also, how can a device that was known before the update suddenly be totally unrecognisable?

Loggare: zigpy.application Källa: runner.py:154 Inträffade först: 07:38:58 (8 tillfällen) Senast loggade: 08:10:56

Unknown device AddrModeAddress(addr_mode=<AddrMode.NWK: 2>, address=0x0B3E)

[dinuxlt]

same here

It was working for >1 year, and now for last week (from 2025.4 (?)) it's not usable. One minute I can't control lights (getting Failed to send request: Failed to deliver message: <sl_Status.ZIGBEE_DELIVERY_FAILED: 3074>) second minute it works as expected. After some time (let's say 10mins) again, not working. All automations which controls the lighting, fails randomly with the same Failed to deliver message errors.

Using SLZB-06M in ZHA mode. Tested old and new firmwares, no different. Can't rollback to 2025.3, I didn't create a backup :(

For light control using Tuya and Moes modules (both with negative wire) both has issues

Totally disabled wifi, no difference.

Noticed, that if I go to Zigbee integration, Zigbee device, Menage Zigbee Device and requesting app_version (or any other field) I get the answer and device starts working for some time.

[Esp32-zapper]

I have now taken away all new hardware from the network, that was installed after this phenomenon started. Nothing changed. Seems like Zigbee routing is not working correctly or routers refuse to use mesh, only communicates with the once they was originally paired with. It affects routers much more then enddevices, in my network.

[Esp32-zapper]

A new firmware for SkyConnect this morning. Would say it's a bit more stable now but network virtualization map still looks funny. Things 3 rooms away seems to have better reception then one just a meter away. Some devices only have dotted lines even if they have strong routers nearby.

[chmac]

I also have this issue, happy to find a bug report already, I was writing about it on the forum. One thing I've noticed is that if I reload the Zigbee service, then some of my problem devices start to work again for a minute or two. I've had to do that to switch some lights on and off. For a while I thought the problem was that one device had been powered off for a while, and powering it back on seemed to fix things, but this morning a similar set of devices are offline again.

[chmac]

I looked through all the related tickets here and these might be a similar issue #136420, #134295, #127786, and #126364.

It seems like there are a few people who have similar issues from time to time. Sadly I couldn't find any consistent fix in any of the threads. It seems like powering off the whole network for a while can sometimes help. Also adding new devices should be approached with caution. This comment has some useful tips.

I was hoping this issue had been caused by a recent update and would be fixed by an update, but by now I think there's probably a problem in my network somewhere and I'll need to spend time trying to resolve it rather than waiting and hoping.

[Auka84]

I also have the same since the update to 2025.4

Many of this error when I try to control any zigbee device

[JeremyGree]

Hi Guys, not sure if this helps but I find that after every reset of the HA (eg after updates) I lose peripheral devices. After 2 or 3 days they come back. It feels like the system is forgetting the topology and taking a while to refind it??? By peripheral devices, I mean the lights in my workshop outside. I had a look though the ZHA UI to see if there was an option to "reload" the previous topology but couldn't find it. I'm only a humble hardware Engineer! Jeremy.

[Esp32-zapper]

Hi Guys, not sure if this helps but I find that after every reset of the HA (eg after updates) I lose peripheral devices. After 2 or 3 days they come back. It feels like the system is forgetting the topology and taking a while to refind it??? By peripheral devices, I mean the lights in my workshop outside. I had a look though the ZHA UI to see if there was an option to "reload" the previous topology but couldn't find it. I'm only a humble hardware Engineer! Jeremy.

Sound more like what's happening if you change Zigbee channel. Not all devices will understand a channel change directly, typically enddevices that was asleep when the change occured. Have you, by any chance, enabled the automatic channel selection to counter a wifi that keep changing channel?

[JeremyGree]

Hi Guys, not sure if this helps but I find that after every reset of the HA (eg after updates) I lose peripheral devices. After 2 or 3 days they come back. It feels like the system is forgetting the topology and taking a while to refind it??? By peripheral devices, I mean the lights in my workshop outside. I had a look though the ZHA UI to see if there was an option to "reload" the previous topology but couldn't find it. I'm only a humble hardware Engineer! Jeremy.

Sound more like what's happening if you change Zigbee channel. Not all devices will understand a channel change directly, typically enddevices that was asleep when the change occured. Have you, by any chance, enabled the automatic channel selection to counter a wifi that keep changing channel?

Thanks for the quick response; that's an interesting suggestion! I've checked but it's set to channel 25. Jeremy.

[Esp32-zapper]

Hi Guys, not sure if this helps but I find that after every reset of the HA (eg after updates) I lose peripheral devices. After 2 or 3 days they come back. It feels like the system is forgetting the topology and taking a while to refind it??? By peripheral devices, I mean the lights in my workshop outside. I had a look though the ZHA UI to see if there was an option to "reload" the previous topology but couldn't find it. I'm only a humble hardware Engineer! Jeremy.

Sound more like what's happening if you change Zigbee channel. Not all devices will understand a channel change directly, typically enddevices that was asleep when the change occured. Have you, by any chance, enabled the automatic channel selection to counter a wifi that keep changing channel?

Thanks for the quick response; that's an interesting suggestion! I've checked but it's set to channel 25. Jeremy.

AND you fixed the wifi channel in you router to something that doesn't collide?

[JeremyGree]

Hi Guys, not sure if this helps but I find that after every reset of the HA (eg after updates) I lose peripheral devices. After 2 or 3 days they come back. It feels like the system is forgetting the topology and taking a while to refind it??? By peripheral devices, I mean the lights in my workshop outside. I had a look though the ZHA UI to see if there was an option to "reload" the previous topology but couldn't find it. I'm only a humble hardware Engineer! Jeremy.

Sound more like what's happening if you change Zigbee channel. Not all devices will understand a channel change directly, typically enddevices that was asleep when the change occured. Have you, by any chance, enabled the automatic channel selection to counter a wifi that keep changing channel?

Thanks for the quick response; that's an interesting suggestion! I've checked but it's set to channel 25. Jeremy.

AND you fixed the wifi channel in you router to something that doesn't collide?

I did. It's as far away as possible -as are my neighbours!

[fmc-fmc]

Thanks for this issue, I'm so glad I've found it. I've been relocating my hub, disabling 2.4Ghz WIFI, added new automations to retry issuing zigbee commands but in the grand scheme of things everything just gradually got worse and worse (while sometimes occasionally working for a couple of minutes).

Looking at this issue I reverted to a backup with 2025.3.3, and so far everything in ZHA just works.

For me it started with a lot of "delivery failed 3074", but lately I just got "zha.issue_zigbee_cluster_command unknown error"-s before reverting to 2025.3.3. No new devices were added between 2025.3.3 and 2025.4.1 in my setup.

[Esp32-zapper]

Thanks for this issue, I'm so glad I've found it. I've been relocating my hub, disabling 2.4Ghz WIFI, added new automations to retry issuing zigbee commands but in the grand scheme of things everything just gradually got worse and worse (while sometimes occasionally working for a couple of minutes).

Looking at this issue I reverted to a backup with 2025.3.3, and so far everything in ZHA just works.

For me it started with a lot of "delivery failed 3074", but lately I just got "zha.issue_zigbee_cluster_command unknown error"-s before reverting to 2025.3.3. No new devices were added between 2025.3.3 and 2025.4.1 in my setup.

What kind of a coordinator do you use? Why I ask is because all problems just disappeared after yesterday's upgrade of the SkyConnect firmware. I use SkyConnect with latest firmware and 2025.4.1 and it sems to be working fine. No more delivery errors. Haven't tried to re-add all the devices that I removed yet, but so far so good. Ordered a 2nd coordinator to convert to z2m and if problems returns, that will absolutely be the way to go.

[mattcee233]

I have been having similar issues - pretty much all of my devices have gone Unavailable since I upgraded yesterday... - not sure if this is the 2025.4 update or if it's the SkyConnect firmware update

Running on a HA Yellow if it helps... everything was solid before I updated yesterday

[fmc-fmc]

Thanks for this issue, I'm so glad I've found it. I've been relocating my hub, disabling 2.4Ghz WIFI, added new automations to retry issuing zigbee commands but in the grand scheme of things everything just gradually got worse and worse (while sometimes occasionally working for a couple of minutes). Looking at this issue I reverted to a backup with 2025.3.3, and so far everything in ZHA just works. For me it started with a lot of "delivery failed 3074", but lately I just got "zha.issue_zigbee_cluster_command unknown error"-s before reverting to 2025.3.3. No new devices were added between 2025.3.3 and 2025.4.1 in my setup.

What kind of a coordinator do you use? Why I ask is because all problems just disappeared after yesterday's upgrade of the SkyConnect firmware. I use SkyConnect with latest firmware and 2025.4.1 and it sems to be working fine. No more delivery errors. Haven't tried to re-add all the devices that I removed yet, but so far so good. Ordered a 2nd coordinator to convert to z2m and if problems returns, that will absolutely be the way to go.

SONOFF Zigbee 3.0 USB Dongle Plus V2 Running on Hardkernel ODROID-M1

Before performing the rollback to the (not even so) old backup, I was on latest greatest HA / HAOS this morning. My mesh consists of the coordinator, 15 mains powered routers (13x blind control, 2x IKEA Tradfri light), 1 battery powered remote. Somemains powered routers suddenly went unavailable (red) in the network view and still could receive commands, while some others displayed as available (green) could no longer be reached (delivery failed 3074 ---> unknown error). It was super frustrating, especially as I was certain the issue was on my end. :-)

Now everything is both green in the network view, and available in reality too (as it should).

[dinuxlt]

Using SLZB-06M coordinator. Up/downgraded firmware, swapped power/LAN cables, power supply, zigbee channels, wifi channels, disabled all wifi access points, changed coordinator location.. all network shutdown for ~1hour.. (leaving devices on) nothing helped.

As result migrated from ZHA to Z2M (Zigbee2Mqtt) and it works now super fast and responsive. Took some time, because I have 45 zigbee devices and a lot of automations. But, as I said, it works now with no issues.

After this my windows blinds motors (Moes) reacting instantly. Before it was lazy, reacting time, some times was few seconds, some times 10-15 seconds.

[puddly]

@mattcee233 Please open a separate issue and attach the ZHA integration diagnostics in addition to a debug log of ZHA starting up (you can enable debug logging, reload the integration, and let it run for 10 minutes or so).

[Auka84]

Using SLZB-06M coordinator. Up/downgraded firmware, swapped power/LAN cables, power supply, zigbee channels, wifi channels, disabled all wifi access points, changed coordinator location.. all network shutdown for ~1hour.. (leaving devices on) nothing helped.

As result migrated from ZHA to Z2M (Zigbee2Mqtt) and it works now super fast and responsive. Took some time, because I have 45 zigbee devices and a lot of automations. But, as I said, it works now with no issues.

After this my windows blinds motors (Moes) reacting instantly. Before it was lazy, reacting time, some times was few seconds, some times 10-15 seconds.

Hey. Is it easy to migrate to Z2M?

I have 70 devices and about 15 has issues with this 3074 error code

[chmac]

Hey. Is it easy to migrate to Z2M?

I have 70 devices and about 15 has issues with this 3074 error code

I think the whole network has to be setup again, and all devices repaired. Sounds like a lot of work. I just ordered 3 SLZB-06 PoE coordinators and I will try splitting my network into sections with multiple zigbee2mqtt installations, hoping that improves stability. I read something about being able to update home assistant without updating zigbee2mqtt which sounds like a nice advantage. If it ain't broke, I won't have to fix it. :-)

[Auka84]

@chmac If Z2M is working better than ZHM, it's not a problem to add all of my device again but i mean, I just need to deactivate ZHM, then install mosquitto broker and then Zigbee2MQTT ? is it all ?

ZHM was really easy to install and to configure, so it was my first choice when i created my zigbee network :p

thx

[dinuxlt]

Yes, you need setup everything from beginning. And all devices should be paired as new (and there was some challenges here).

[chmac]

@chmac If Z2M is working better than ZHM, it's not a problem to add all of my device again but i mean, I just need to deactivate ZHM, then install mosquitto broker and then Zigbee2MQTT ? is it all ?

I believe that's correct, I haven't done it yet. I guess some, maybe most, devices will need to be factory reset which can be a real hassle. Some bulbs for example take a lot of trial and error to reset. I'm hoping to move some problem devices first and hope that the two networks can coexist on different channels while I slowly migrate stuff and decide if I want multiple networks or not, etc.

Edit: I think there's also no guarantee that zigbee2mqtt is more stable or reliable than ZHA! But it's possible to not update it if it's working, while still updating home assistant, and possible to run multiple networks simultaneously.

[Auka84]

@chmac The other solution I didn't try is to change the channel. But I guess i also need to repair all device And some devices are difficult to access (behind washing machine for example) btw i notice than all devices with issues are old zigbee devices, or led strip bought on aliexpress. I haven't any issue with my aqara zigbee switches, compatible with zigbee 3.0 for example

[Esp32-zapper]

@TheJulianJES I think I got this now. It wasn't 2025.4 that was the problem. This had been going on since 2025.3 or maybe even earlier. Every update after that made it just a bit more problematic. This problem is (I think) how some routers are handled. I have hade some zigbee led light strips for some time. They were fine in the beginning but it's them that now messes up the mesh network. I removed all devices in my network that I installed during the last couple of months and after the network had balanced itself again (+ maybe the latest SkyConnect firmware), the network was rock solid again.

The plan was to reinstall the devices I removed one by one with some days apart to see if any of the was the problem.

I now tried today to reinstall the first one and it didn't take long before the 1st problem started again, with random devices failing with <sl_Status.ZIGBEE_DELIVERY_FAILED: 3074>. Removed the device and after a few restarts everything is rock solid again.

The led strips is a ZB-DL01 by eWeLight (also versions for rgb and rgbw and rgbcct. Thing is, when I got them, everything worked great, both the strip and the rest of the Zigbee network but now it seems like all messages routed through one of them will disappear.

[Auka84]

@Esp32-zapper Oh right ? I also have this led strip bought on aliexpress and another ewelink device

So without this led strip, all is working fine ?

[Esp32-zapper]

@Esp32-zapper Oh right ? I also have this led strip bought on aliexpress and another ewelink device

So without this led strip, all is working fine ?

According to my test

-Removing them all made the network rock solid again, after the network had rebalanced itself and removed those routers from the "routing table". -Adding just one of them again, it didn't take long before the first problems started. -removing it again and the network self healed again.

Would be great if you could do the same test.

Just beware, you have to plug them out as well or those f.ckers will auto join again.

[Auka84]

@Esp32-zapper I just deleted the both. I'll see if it's better. I'll wait some hours now :)

[Esp32-zapper]

@Esp32-zapper I just deleted the both. I'll see if it's better. I'll wait some hours now :)

Do a few restarts during that time to help the network to clean up