I have been working for weeks trying improve my ZigBee network. I had the following symptoms:
• Lossing lines to endpoint nodes
• Red Flag Message “Failed to execute LQI for Coordinator” when requesting map
• Devices dropping offline
• Zigbee related automations intermittently failed to execute
I searched this forum and the internet for solutions. I did the following to no avail.
• Updated the coordinator firmware. (SONOFF Zigbee 3.0 USB Dongle Plus (CC2652P)
• I added smart plugs routers/repeaters to the network
• Repaired most devices several times
• Deleted device from ZigBee network then repaired several times
I was getting nowhere fast. With the resent 2023.11.0 update I lost my Moen Flo integration. Because I run HA on a VirtualBox 6.x VM, I was able to go back to a 2023.10.x snapshot and prove to myself 2023.11.0 killed my Moen Flo integration. I reported the problem on GitHub. I was latter informed the Moen Flo integration would be fixed in 2023.11.2.
I was getting quite frustrated. I decided to bite the bullet and rebuild my system from start. Not really from start, I used a backup.
First, I upgraded VirtualBox to version 7.x. I thought I was installing a separate VirtualBox by shutting down VB 6 and choosing a different path for the install. I’m no computer genius but I get by. In the end of the VB 7 install, I saw my older HA snapshots, but they would not run. I reality I did a VB 6 to 7 upgrade. Ok
I followed and adapted the directions I found on the internet to build a new Home Assistant VM. I downloaded and used a new (.vdi) for the new VM.
I started up HA and used my most resent backup to configure HA. The restore seemed to never end. I even went on an hour plus errand. When I returned the onboarding was still thrashing away. I hit the back button and Logged in. Everything seemed to come back. I was a little amazed. Moen Flo was still dead but the fixed was coming in 2023.11.2. I recall only having to reconfigure my network storage location.
The system seems snappier. I monitored my ZigBee network for several days. It was most definitely more robust. End nodes stopped dropping offline. But when I selected to see the map, I still got the red flag message “Failed to execute LQI for Coordinator” and the map was still not showing most lines to the endpoint nodes.
I found the solution by search the internet for “zibee2mqtt no line to endpoint”. The search returned the following link to GitHub.
The solution for the fix was to correct the coordinator network address. Open the File Editor and navigate to /config/zigbee2mqtt/database.db. Line 1 in the (.db) file will have the network address or “nwkAddr”. My address was 54084. It should be 0 per the threads in GitHub dialog in the above link. I stopped Zigbee2Mqtt and edited line 1 of the file so the coordinator network address was 0 (zero).
……“nwkAddr”:0,“manufId”……
I started Zigbee2mqtt. Problem Solved