Z2M ENOTFOUND failures

My Z2M addon is frequently stopping and the logs indicate it is because it is unable to connect to my zigbee controller:

[2026-02-13 16:27:50] e[32minfoe[39m: 	z2m: Logging to console, file (filename: log.log)
[2026-02-13 16:27:50] e[32minfoe[39m: 	z2m: Starting Zigbee2MQTT version 2.8.0 (commit #unknown)
[2026-02-13 16:27:50] e[32minfoe[39m: 	z2m: Starting zigbee-herdsman (9.0.2)
[2026-02-13 16:27:50] e[32minfoe[39m: 	zh:zstack:znp: Opening TCP socket with slzb-06.internal:6638
[2026-02-13 16:27:50] e[31merrore[39m: 	zh:zstack:znp: Socket error Error: getaddrinfo ENOTFOUND slzb-06.internal
[2026-02-13 16:27:50] e[31merrore[39m: 	z2m: Error while starting zigbee-herdsman
[2026-02-13 16:27:50] e[31merrore[39m: 	z2m: Failed to start zigbee-herdsman
[2026-02-13 16:27:50] e[31merrore[39m: 	z2m: Check https://www.zigbee2mqtt.io/guide/installation/20_zigbee2mqtt-fails-to-start_crashes-runtime.html for possible solutions
[2026-02-13 16:27:50] e[31merrore[39m: 	z2m: Exiting...
[2026-02-13 16:27:50] e[31merrore[39m: 	z2m: Error: Error while opening socket

However, the host itself can resolve the SLZB-06 address just fine. Eg. from the HA CLI i drop into the host using ‘login’ and nslookup gives the SLZB-06 address just fine.

The host is a Home Assistant VM on a machine physically connected to the same switch the SLZB-06 controller is connected to (ie. by ethernet).

I’ve tried specifying the ip address of the SLZB-06 in my Z2M yaml, but I still get ENOTFOUND error. So, it doesn’t seem to be an issue with Z2M being able to resolve an ip address.

I’m a bit stumped as to why Z2M is unable to connect to the SLZB-06. It doesn’t seem to be a name resolution issue since I get the same issue when i specify the ip address, and the host itself has no issue resolving the address. Grateful for any thoughts/suggestsions.

Is it connected thru a switch that has a WIFI connection between the router and the switch. Like a Mesh by WIFI router?
might be a WIFI interruption. (intermittently)

No, the machine that hosts the Home Assistant VM and the SLZB-06 are both connected to the same switch by ethernet. The Home Assistant VM itself has no problem resolving the SLZB-06 address and can ping the SLZB-06. I try and start Z2M add on and I get the ENOTFOUND errors (intermittently unfortunately).

OK, I installed the ‘Ping’ integration and when Zigbee2MQTT goes down, the ping integration is unable to detect the SLZB-06 host:

However, when I ‘login’ from the Home Assistant CLI, the Home Assistant machine itself can ping the SLZB-06 just fine. The SLZB-06 is running just fine and I can access its web interface directly from a laptop.

So, the issue doesn’t appear to be a problem with either the SLZB-06 or with Zigbee2MQTT. It appears to be something about how Home Assistant’s internal networking works. I know Home Assistant uses containers but I don’t understand how networking works with these. Grateful for any insights.

I tried (and ruled out) other solutions such as enabling multithreading on the SLZB-06 as suggested here:

I have created an automation that runs every 5 minutes and attempts to start Zigbee2MQTT if it’s not running. This eventually works. For example, when Zigbee2MQTT last went down, the automation tried restarting it every 5 minutes for 45 minutes before it worked. Whatever the issue with networking is, it eventually self corrects.

Following up in case it is of any use to anyone.

This has been a really annoying issue. All signs pointed to a problem with the Zigbee2MQTT addon (app). The Z2M logs showed lots of failed attempts to connect to the SLZB-06, apparently unable to resolve the address, but there were also various errors suggesting ‘connection reset’. This was suggesting a networking problem.

Checked my network and no problems apparent. I could ping the SLZB-06 OK from various machines on my network. I logged in to HA CLI (using ‘login’) command and I can ping the SLZB-06 OK from the HA (virtual) machine itself.

So I open a shell in the Zigbee2mqtt docker container itself and cannot ping the SLZB-06 from there. So, this suggested some sort of networking error within the Z2M docker container.

I created a HA automation that detects when Z2M goes down and restarts the ‘Core’ and ‘Zigbee’ on the SLZB-06. This worked to a degree but I would still have periods of up to several hours of attempted restarts of Z2M.

Eventually, I found a comment by Koenkk (author of the HA Z2M plugin/app?) on GitHub:

It could also be an issue of the coordinator itself, in the past I read people having these issues when opening the web page of the coordinator.

From the following thread:

Now, I had the SLZB-06 web interface permanently open on my laptop recently while trying to diagnose these issues. The SLZB-06 has an option under ‘security’ to disable its web server when a socket (from Z2M?) is open. I enabled this option and Z2M seems to have been much more stable for the last few days.

To get back in to the SLZB-06 web interface, I now need to stop Z2M on HA. I hope this works longer term.