Introduction

1. What This Module Covers

Every connection between two services can fail, even when the corresponding arrow looks simple in a system diagram. A network call may stop responding, return an error, complete partially, or finish too slowly to remain useful.

These outcomes are part of normal production operation rather than unusual exceptions.

Individual failures may be rare, but a large request volume makes them occur continuously.

At 100,000 orders per day, a failure rate of only 1 in 200 still produces 500 failures every day. A reliable design must handle that measured volume.