Failed to deliver message: <EmberStatus.DELIVERY_FAILED: 102>

Configuration Zigbee
82

Same problem here. Zigbee with HA SkyConnect and ZHA integration. I had som 60+ devices (including frient smoke detectors) and everything worked just fine.

Three weeks ago I added 2 additional smoke detectors and 4 water sensors. Since then (maybe a co-incidence) Zigebee - Network ist almost unusable, I have to reboot the system couple of times a day. After reboot Zigbee works for a couple of hours but it’s really frustrating.

Any solutions or ideas? I already switched to another channel - no difference.

83

I fought with my zigbee network for three weeks.

  • Changed channels (wifi and zigbee)
  • Longer usb cable
  • Added more routers (though I already had 20+ routers)
  • implemented zigbee groups instead of home assistant light areas
  • shut down the coordinator for 40mins to force a zigbee network rebuild
  • tried various firmware

I still had super laggy automation and I’d get zigbee failed to deliver message errors all the time in my logs. Performance was TERRIBLE.

I was using a home assistant green and a skyconnect dongle.

Luckily, my wife and kids were away for a week…

I wanted to try a new coordinator to see if that was the issue. I tried to migrate the radio to a sonoff-E, but that left me with flickering lights and a repeatedly crashing ZHA.

That was the last push I needed…

I reflashed my home assistant green and started COMPLETELY from scratch.

I had a second sonoff-E, so used that (as I wanted to make sure that the partially successful radio migration to the first sonoff-E coordinator didn’t carry over to my fresh start).

I re paired all 62 of the zigbee devices EXACTLY as I had before (mains powered closest to the coordinator first, then worked my way away from the coordinator adding just the mains devices. Then I repeated for battery powered and end devices). The nerwork map essentially looked identical before and after.

I then recreated all of my scenes and automations, sensibly using zigbee groups where I could, but also using the more flexible HA areas.

It all worked… beautifully… And with essentially zero errors in my log file.

  • Same 62 devices
  • Same zigbee channel
  • Same usb extention cable
  • Same home assistant green, sitting in the same server rack
  • Pretty similar automations and scenes (they functionally did the same things, but their YAMLs would undoubtedly been inconsequentially different)

The only thing that was different, was swapping the nabu Casa skyconnect for the sonoff-E. I realize this probably should make zero difference as they use the same radio chip, but I can tell you it is absolutely night and day different.

I was ready to switch back to smartthings (if my wife didn’t force me to first because nothing was working).

Now, it’s back to working perfectly… I’m creating several backups and holding my breath.

Three days and not a single failed to deliver message (and I was previously getting several an hour).

84

Congratulations! Personally, I switched channels for the nteenth time about a month ago and have only received the 102 Ember error one time. Go figure.