From yesterday i have suddenly a problem with my HA, and it appears to be connected with my database.
I have MariaDB addon installed in my HAOS, so all data goes into that (i have recorder set to 15 days). Additionally al history is also copied to Influxdb, which is installed on my Synology.
Yesterday it happened first time. Dashboards in HA work, but that’s about all. No addons can be shown/started (like esphome, studio, grafana…). When i click “check configuration” it says that configuration.yaml is missing. I can’t connect to my HA via ssh, since connection is terminated immediately. So my only option is force-restart (by removing power). When i force-restart HA machine (it’s intel NUC) all is back perfectly. What’s even more odd is that all history IS THERE!
Yesterday it happened exactly at 12:00 (AM), today it happened exactly at 10:00…
Does anyone have a clue what’s going on and where to start looking? I didn’t install anything (that i remember) from yesterday.
I did install a kind of HDD smart monitor yesterday after first error and it says all smart data are ok on my NVMe ssd, so it appears that ssd is not to blame. Could it be RAM?
I have log saved into my Synology and it’s filled with pretty much same data:
ERROR 2024-09-03T12:15:18.572623+02:00 : [‘Error executing query: (MySQLdb.OperationalError) (2005, “Unknown server host 'core-mariadb' (-11)”)\n(Background on this error at: Error Messages — SQLAlchemy 2.0 Documentation)’]< >(‘components/recorder/util.py’, 202)
tons of these lines…i’d says 100 per second.
From this error i’d say that connection to db is broken… but it’s not if data is still coming into db…?