Why does a cooling failure in one datacenter brick the entire ecosystem? Reliability 101 to have redundancy here I'd think...
Not sure if that indicates how much of a backlog of emails Proton has because of this outage, or if the DR infra is just that much slower?
Manually?!