LinkedIn's Deployment Crisis and Why Delivery Matters

The Fundamentals and Future of DevOps and Software Delivery - Yevgeniy Brikman

Topic 4 of 9

Published Feb 17, 2025 · Clip 23:02–32:13 (9:11)

Yevgeniy recounts LinkedIn's 2011 crisis: despite hypergrowth and a successful IPO, a two-week release train deployment catastrophically failed over multiple days, forcing Project Inversion—halting all product development so everyone worked on tooling until deploys became reliable (eventually hundreds of deploys per day; Apache Kafka emerged from that era). He gives three reasons to learn software delivery: avoid getting into that state, gain independence to ship your own ideas, and write better app code by understanding architecture, deployment, and security.