ZHA Dropouts, constant log warnings with Sky Connect

Decided to finally take the plunge and got a new NUC and a Sky Connect dongle and figured i would get everything set up from scratch but slowly migrating everything over from my old setup which consisted of HA running on a Synology Docker and a Smartthings Hub that was used as a devicehandler more or less.

Everything was working fine on the old setup, few times did i get a latency of 300+ milliseconds on the Smartthings integration trough cloud but nothing massive, but there was a few features I was missing due to the lights and sensors going through the Smartthings integration on the device and entities itself.

So anyway, started pairing all the repeaters (lightbulbs) closest to the dongle and then moving outwards, all devices was paired at or close to the final destination to get the best channel mapping. and the channel was selected by checking for all Wifi channels for any interference to minimize any issues.

After pairing about 60 devices i started moving over all the automations and after a few days i started noticing that some lights would not have turned off when they were supposed to, waking up and lights would have been on all night, even turning on a few for no reason, without any automations running.

after a few weeks now and it’s hard to say if it’s been getting worse or not but i have tried deleting the database (1,2 Gb in only 2 weeks). disabling all the integrations, removing all Z-Wave JS devices and dongle, uninstalling all the Add-ons etc. but i keep having issues and i do get these warnings in the log.

i stopped migrating over more devices until i can figure out what the issue is, but right now I´m at a loss. Anyone know what these log events are? and can i get any more debugging information somehow?

FYI I’m using ZHA groups to group all fixtures with multiple bulbs, maybe 20 groups with 2-4 devices in each. the rest are all individually managed.

2023-01-23 18:11:29.562 WARNING (MainThread) [homeassistant.components.zha.core.channels.base] [0xC6C5:1:0x0300]: async_initialize: all attempts have failed: [DeliveryError('Failed to deliver message: <EmberStatus.DELIVERY_FAILED: 102>'), DeliveryError('Failed to deliver message: <EmberStatus.DELIVERY_FAILED: 102>'), DeliveryError('Failed to deliver message: <EmberStatus.DELIVERY_FAILED: 102>'), DeliveryError('Failed to deliver message: <EmberStatus.DELIVERY_FAILED: 102>')]

these keep getting repeated with different IDs, one or two matches the ID of some ZHA Groups i have but most don’t and so I´m not sure if there is any connection or not sadly.

My map,

Any tips are welcome, need to resolve this so i can migrate all the rest of the devices, automations, integrations, dashboards etc.
Or i might have to consider going back to Smartthings altogether.

Edit, Almost forgot to mention that all the devices has had its OTA firmware updated (IKEA and ledvance) and 99% of the devices are Ikea light bulbs, Philips Hue Motion sensors (indoor and outdoor) and Smartthings Door sensors. removed everything else so far to keep it clean.

I have no idea what could cause such symptoms other than automations or groups causing issues, or maybe the devices with issues were not fully factory reset before pairing or something so probably a good idea to fully factory reset and re-pair each device that has issues(?). Also post the debug logs.

https://www.home-assistant.io/integrations/zha#debug-logging

Anyway I’ve collected general tips about removing interference which can cause intermittent issues:

https://community.home-assistant.io/t/guide-for-zigbee-interference-avoidance-and-network-range-coverage-optimization/515752

Are you use a USB extension cable - either the one that came with it or a longer one?

Sorry for the very late reply and not leaving a responce.

Appreciate the link @Hedda, havent had time to go through all of it just yet but it looks extensive so i will hopefully find something i havent thought about. On top of all these issues my ISP has DNS problems and my domain is blocket so external adress is dropping causing another set of errors in the log.

Also figured out that some of my issues are due to a limitation of the Zigbee protocall, what i understand is a broadcast (only real Zigbee Groups are affected) that limits it to 8 groups per 9 seconds and i rely on quite a few groups in my new setup to ensure lights are synced and easy of use in the automations.

Automations issue/ ZHA Network busy errors after migrating to Skyconnect dongle · Issue #86411 · home-assistant/core (github.com)

@ha_steve i Actually started using a longer cable (3 meters) insteaad of the supplied extension but cant say i noticed any real world difference and the map looks identical to before.

Once i get the time to fix all the Zigbee groups and broadcasts (and all automations that uses them) it will hopefully solve most of the issues I´m having, will keep the thread updated.