AWS Wide Spread Outage

Examine the December 7, 2021 US-EAST-1 incident as a control-plane and dependency failure. Distinguish an application that is still running from its ability to launch replacements, change configuration, authenticate, or observe its own health.

Introduction

On December 7, 2021, congestion between AWS’s internal and main networks impaired several services in Northern Virginia. Recovery times differed by service. AWS’s incident report describes a regional event with broad downstream effects, rather than a failure of every AWS region or every running workload. The useful design question is which shared operation made an otherwise healthy application unavailable.

The following sequence describes the chain of technical events that led to the outage.

Sequence of events