Hi,
I have OH3.1 running on an Rasp4 booting from SSD, incl. Mosquitto from openhabian-config. Everything is up and running. Problem: in my log there seems to be a second MQTT Broker running. This ends up in these two entries every minute:
2021-10-17 21:37:00.226 [INFO ] [.reconnect.PeriodicReconnectStrategy] - Try to restore connection to 'localhost'. Next attempt in 60000ms
2021-10-17 21:37:00.229 [INFO ] [.transport.mqtt.MqttBrokerConnection] - Starting MQTT broker connection to 'localhost' with clientid 6a8d9c85-d4f6-49e7-a6ae-f59c482b8819
In “things” there is only one MQTT Broker, having this clientid: 15e1fac0-2c73-4987-bd42-ad0687a619a8 (->thing → code)
I assume that during my hazzle with getting MQTT running I did sth wrong.
If this is an embedded or local broker, this seems to be totally ok. The client is the openHAB MQTT binding connecting as a client to the local broker.
I see what you mean. The broker has a clientid property but as it is the broker, it does not need to connect as a client to himself or another broker.
Once again, seems to be totally ok. Besides that the client (openHAB) cannot establish a connection. Can you try to connect to the broker from the machine itself with any MQTT client available?
There are NO brokers hidden in openHAB. A broker is an external service of some kind (usually Mosquitto running on the same host of course).
Check your log carefully, all attempted connection attempts to broker(s) will be logged, so if you think this is an “extra” one then your regular one will also be there somewhere.
Make sure you have not configured a “SystemBroker” bridge thing, this is redundant in OH3 and never needed.
Hi Frank,
thank you so much for the input so far! I was struggling with my log getting flooded as this is a new pattern to me. I had 2.5 up and running (inkl. MQTT) and there was no such behaviour.
Where could I set this up? (Sorry, I fear I don’t know exactly what you mean).
