Hi everyone, I need some help with a configuration.yaml issue!
At first, my system was working fine, and the battery status (along with other sensor data) was displaying correctly in Home Assistant. However, we installed MariaDB to extract the data, and for this, we made some changes to the configuration.yaml file as we have to this case.
After making these changes, the battery information stopped displaying correctly on the Home Assistant interface. We then reverted to the default configuration.yaml file and ran it again, but now the data is still not showing properly on the main screen.
Has anyone faced a similar issue? Could there be something still affecting the data display, even after reverting to the default configuration? Any guidance would be greatly appreciated!
The screenshot of Logs contains several significant errors and none of them appear to be related to the change to MariaDB.
The first error in the list has been logged over 13000 times and is to due to script.publish_keepalive doesn’t exist.
The second and third errors are caused by the occasional inability of the Met.no integration to contact the Met.no servers.
The fourth error appears to be due to some sort of illuminance sensor attempting to connect to your Home Assistant system using the wrong login credentials.
The fifth error appears to be due to the received data not being in JSON format, thereby causing the template to fail.
The last two errors appear to be related and due to a casting failure caused by some sort of networking error.
You should reserve some time to fix these errors individually (where possible). Otherwise they can all accumulate to the extent that no one knows where to start to help you fix the many failure symptoms they produce.
Go to Developer Tools → States and find sensor.load_ca_power and sensor.batt_voltage. What are their reported state values?
In addition, what is each sensor’s underlying integration? In other words, which integration produces those sensors?
The errors in the dashboard indicate the state of the entity is unknown. That indicates that an integration (the one delivering the entity) isn’t working. That is not related to the recorder, so it is no indication if the default sqllite or mariadb is working. The integration page for the integration may indicate a problem loading, there likely also is something in the logs about the cause.
It is possible the integration providing the entities that are unavailable was set up via configuration.yaml and that you accidentally did something to it. It is also possible that the reboot was the start of the problems, and MariaDB just the time that the reboot was done. But it is impossible to tell based on the little info you gave.
