DevOps Daily with Fexingo · 2026-07-23 · 10 min
Episode 129 of DevOps Daily with Fexingo dives into a subtle but dangerous Kubernetes behavior: PodDisruptionBudget violations during voluntary evictions. Lucas and Luna walk through a real incident where a cluster running a critical payment service on 3 replicas with minAvailable=2 saw a complete outage during a node drain. The problem? The PDB only counts pods that are healthy according to readiness probes - but the operator assumed all 3 were eligible. When one pod had a stale ready state and another was being recreated, the eviction controller allowed the drain to proceed, dropping below the minimum. The hosts explain how PDBs interact with readiness gates, why 'voluntary disruption' is misleading, and a concrete fix: use maxUnavailable instead of minAvailable, and always combine with PodHealthyPolicy. No hot takes - just an operational war story that could save your next maintenance window.
How Kubernetes ServiceAccount Token Expiration Breaks CI Workflows75 / 100How Kubernetes StatefulSet PVC Resizing Causes Node Disk Failures94 / 100How Kubernetes Topology Spread Constraints Create Scheduling Hotspots95 / 100How Kubernetes CRD Versioning Breaks Controller Upgrades90 / 100How Kubernetes Audit Logging Causes etcd Performance Degradation91 / 100