Hacker News Reader

>We are investigating connectivity issues affecting some instances in a single Availability Zone in the US-EAST-1 Region.

Well there’s your problem, people. Use multiple AZs.

Easier said then done if that would mean synchronizing database and filesystem that is heavily written to.

Depends on what you use. RDS can span AZs and failover in events like this.

AWS always understates the scope on the status page.