last sunday my DB got corrupted. (i suspect the recorder repack running into ENOSPACE).
HASS just renamed it and started a new DB.
when i noticed that (due to the lack of a history) half a day later and repaired my DB
i had to decide on sticking with the new DB and lose my history, or just lose the data of the current day.
i understand there’s a trade-off between availability and consistency.
a notification would have at least helped me to notice and act faster, limiting the lost data.
I had a similar issue recently with an external Database (and saving 365 Days recorder-Data )
While a “broken” DB could maybe be monitored (local sqlite file) - such an notification might be difficult on an external system.
but you can try to create your own montioring by using SQL integration - and then, based on DB Size, or SQL Return value, you could create an alarm notification via automation?
I think, this would also allow you to define certain thresholds YOU want to be notified about.
)