Designing distributed systems: Patterns and practices for reliable software systems

Description

Nearly every application now built is a distributed system, and these systems are expected to be reliable, dynamically updatable, and scalable to any load. However, though thousands of distributed systems are activated every day, designing and building them is more black art than science. The good news is that the study of such systems reveals a collection of repeated patterns and practices that can be applied to quickly construct reliable systems.

Brendan Burns describes these patterns and explains how they can be used with the Kubernetes container orchestrator. These patterns—from the simple, like replicated sharded systems, to the more complex, like sharding, functions as a service, scatter-gather, and more—will enable you to build systems more quickly and more easily discuss the systems you have built with collaborators. You’ll also be able to more rapidly bring novice programmers up to speed, without requiring that they learn from experience.