Actually, looks like it was something between their monitoring system & the DB. Per my prometheus stats, DB kept chugging along the whole time.