Exercise: Walk a Regional Failure

1. The Setup, and the Claim Being Tested

A payments product runs active-passive across two regions. The documented position is specific.

RPO   60 seconds        asynchronous replication to the secondary
RTO   15 minutes        documented, reviewed, approved
last drill              component failover in staging, four months ago
on call                 one person, rotating weekly
routing                 DNS, TTL 60 seconds
secondary               full capacity provisioned, serving no traffic

At 02:14 the primary region begins failing