Kubernetes Workload Reliability
44 incident problems about Kubernetes Workload Reliability. Start with the reviewed ones.
먼저 읽을 가이드
추천 문제
All problems (44)
K8S-336A custom metric adapter reports current values while the HPA history window is still full of high samples and scale-down never happens when expected during a failover rehearsalThe metric is honest now, but the autoscaling controller still reasons over a past you forgot to inspect. Normal traffic masked the issue until the standby or alternate path became active under rehearsal conditions.KubernetesIntermediate17 minProK8S-288A PodDisruptionBudget is satisfied in aggregate while one topology domain has no remaining healthy replicas for a targeted drain during a failover rehearsalThe global count looks safe, but the local maintenance blast radius is not. Normal traffic masked the issue until the standby or alternate path became active under rehearsal conditions.KubernetesIntermediate17 minProK8S-127A very generous startup probe masks a crash loop during rollout, and autoscaling expands the broken ReplicaSet before anyone noticesThe app never becomes truly healthy, but the rollout budget is consumed on pods that are still within the startup grace window.KubernetesIntermediate17 minProK8S-240An HPA reads CPU from the main container while the injected sidecar consumes most of the real workload headroom during a failover rehearsalThe autoscaler sees one process and the pod's real bottleneck lives somewhere else. Normal traffic masked the issue until the standby or alternate path became active under rehearsal conditions.KubernetesIntermediate18 minPro