Building the Perfect Infrastructure with Kubernetes
summary
Describes migrating about 20 services in two months onto Kubernetes to manage a 30+ microservice landscape, and building CI/CD including a Docker registry, Jenkins pipelines, feature namespaces and a feature-namespace preview button for testing. Uses Prometheus monitoring, centralized logging and autoscaling to reduce toil and wasted resources, but encountered fourfold cost spikes on initial clusters, Redis cluster and DNS negative-cache failures, slow container I/O (fixed with tmpfs), and required external consultants and six months of iteration to reach stability.
AI-written summary · about
chapters (7)
- 0:00Introduction
- 3:57Old infrastructure problems
- 9:59Choosing Kubernetes
- 15:20First attempt
- 27:25Production issues and fixes
- 33:08Current improvements
- 36:20Migration summary
related talks
20mKeynote: The Challenges of Migrating 150+ Microservices to Kubernetes
35mPerforming Infrastructure Migrations at Scale
45mLearning How to Pronounce Kubernetes to Production in 3 Months!
14mMigrating and running production services in Kubernetes
20mThe path to self-healing: Re-architecting for massive scale on kubernetes
46mAdvanced Kubernetes: Lesson Learned From Building..
46mAround & After Kubernetes: The Principles and Ideas that Guide Us
36mOptimizing Resource Usage in Kubernetes