How Spotify Migrated Millions of Pods Without Downtime | Zero Downtime Kubernetes Migration

Video Description

In this video, Spotify engineers Nick Rutigliano and Daniel de Repentigny share the incredible story of how they re-created Spotify’s entire Kubernetes backend without a single second of downtime. Learn how they migrated millions of pods across tens of thousands of nodes, handling complex orchestration and automation challenges. Discover the key architectural strategies and performance engineering principles they applied to keep the music playing for over 500 million users. If you’re a performance tester or SRE, don’t miss this deep dive into Spotify’s zero-downtime migration journey! - Follow Us on Facebook: https://www.facebook.com/perfology - Connect on Instagram: https://www.instagram.com/perfologys/ - Network on LinkedIn: https://www.linkedin.com/in/perfology/ #performancetesting #sre #kubernetes #devops #cloudmigration #spotify #zerodowntime #automation #backend #perfology This video details how Spotify re-created its entire Kubernetes backend without downtime, migrating millions of pods across tens of thousands of nodes. The speakers share architectural strategies and performance engineering principles that ensured continuous service for over 500 million users. Here are the chapter-wise timestamps: Introduction to Spotify's Platform and Scale (0:00-2:04) Spotify's Backend Architecture and Challenges (2:04-3:01) Kubernetes at Spotify Today (3:01-6:06) Rebuilding Clusters: Product Perspective (6:06-11:51) Cluster Offerings and Key Considerations (11:51-15:19) Architectural Strategy for Zero-Downtime Migration (15:19-17:16) Preparing Clusters and Workloads (17:16-21:28) Migration Tooling and Principles (21:28-25:22) Q&A (25:22-41:25)