For a little over a month, I moved my MQTT to an external docker.
I’m currently on 2.0.20 (pinned) instead of Gladys’ which is on 2.0.15, and I should note that everything works perfectly ![]()
In the last few days there were two docker image changes for mosquitto to 2.0.20 (automatic update with watchtower), I don’t know why (didn’t look at the changelogs) but my Gladys stopped communicating with it.
Looking at the MQTT integration configuration in Gladys, there was no communication.
A simple Sauvegarder restored everything but it took several hours for me to notice (thanks to the tuto Tempo which shows the current cost of my electricity on my dashboard).
I should point out that Z2M had no reconnection issues with the MQTT server after the update.
@pierre-gilles do you think it’s possible to set up a heartbeat to the MQTT server (only for external or internal as well) and trigger the configuration save to automatically reestablish the connection?
Extrapolating a bit on the monitoring system, we could monitor z2m, node-red, netatmo, etc., and trigger an automatic reconnection with a timeout and a message if there’s no reconnection after the timeout?