zlacker

>The plan was always "in the unlikely event that both servers die at the same time, be able to spin HN up on AWS.

>We had done dry runs of this in the past,

Incredible. Actual disaster recovery.