Building Resilient Services on Kubernetes
summary
Intuit runs 7,000 applications across 315 clusters and uses Kubernetes patterns that actually eliminate infrastructure 5xxs: lifecycle-aware Pods, ALB Readiness gates to avoid premature target removal, and ALB slow‑start or synthetic pre-warm requests to avoid cold Java startups. Production config examples include a preStop sleep with terminationGracePeriodSeconds: 62 to cover target deregistration, ensuring the process is PID 1 and handles SIGTERM, plus PodDisruptionBudgets to limit voluntary evictions.
AI-written summary · about
chapters (10)
- 0:00Resilient Kubernetes services intro
- 1:29Intuit’s Kubernetes at scale
- 2:12Client resiliency patterns
- 5:47Pod lifecycle basics
- 8:15Readiness gates and load balancing
- 13:19Graceful pod terminations
- 25:36Introducing new pods without disruption
- 29:11Pod disruption budgets
- 34:50Key takeaways
- 36:16Q&A: Pre-stop hooks
related talks
26mBuilding Resilient Services on Kubernetes
17mResiliency Patterns in Kubernetes
36mOptimizing Resource Usage in Kubernetes
17mResilient Applications on Kubernetes
29mKubernetes Workload Resiliency in Action: Beyond Basics
33mBest Practices for Applications on Kubernetes for Both Frequent Updates & Stabil...
36mResilient Multi-Cloud Strategies: Harnessing Kubernetes, Cluster API, and Cell-Based Architecture
51mBuilding a Resilient Application Infrastructure on Kubernetes