The issue was docker-related and I spent an embarrassingly long time trying to restore the broken db afterwards. No luck.
Good news is I’ve now got automated nightly backups running and it was my own personal instance…so the only person losing data was me.