Introduction

1. What This Module Covers

Most production outages begin with an internal change rather than failed hardware, unexpected traffic, or an attacker. A deployment, configuration update, or migration changes a system that was operating correctly.

The reliability module explained how to survive external failures. This module explains how to release your own changes safely and observe production while those changes take effect.

Five patterns perform two jobs. The first two observe whether a server should receive traffic and where one request spent time across thirty services